We discussed earlier in Part 1 of Blockchain Applications of Data Science on this blog how the world could be made to become much more profitable for not just a select set of the super-rich but also to the common man, to anyone who participates in creating a digitally trackable product. We discussed how large scale adoption of cryptocurrencies and blockchain technology worldwide could herald a change in the economic demography of the world that could last for generations to come. In this article, we discuss how AI and data science can be used to tackle one of the most pressing questions of the blockchain revolution – how to model the future price of the Bitcoin cryptocurrency for trading for massive profit.
A Detour
But first, we take a short detour to explore another aspect of cryptocurrency that is not commonly talked about. Looking at the state of the world right now, it should be discussed more and I feel compelled to share this information with you before we skip to the juicy part about cryptocurrency price forecasting.
The Environmental Impact of Cryptocurrency Mining
Now, two fundamental assumptions. I assume you’ve read Part 1, which contained a link to a visual guide of how cryptocurrencies work. In case you missed the latter, here’s a link for you to check again.
The following articles speak about the impact of cryptocurrency mining on the environment. Read at least one partially at the very least so that you will understand as we progress with this article:
So cryptocurrency mining involves a huge wastage of computational resources, energy, and enough electrical power to run an entire country. This is mainly due to the model of the Proof-of-Work PoW mining system used by Bitcoin. For more, see the following article..
In PoW mining, miners compete against each other in a desperate race to see who can find the solution to a mathematical hashing problem the quickest. And in every race, only one miner is rewarded with the Bitcoin value.
Ethereum goes Green! (From Pixabay)
In a significant step forward, Vitalin Buterik’s Ethereum cryptocurrency has shifted to Proof-of-Stake based (PoS) mining system. This makes the mining process significantly less energy intensive than PoW. Some claim the energy savings may be 99.9% more efficient than PoW. Whatever the statistics may be, a PoS based mining process is a big step forward and may completely change the way the environmentalists feel about cryptocurrencies.
So by shifting to PoS mining we can save a huge amount of energy. That is a caveat you need to remember and be aware about because Bitcoin uses PoW mining only. It would be a dream come true for an environmentalist if Bitcoin could shift to PoS mining. Let’s hope and pray that it happens.
Now back to our main topic.
Use AI and Data Science to Predict Future Prices of Cryptocurrency – Including the Burst of the Bitcoin Bubble
What is a blockchain? A distributed database that is decentralized and has no central point of control. As on Feb 2018, the Bitcoin blockchain on a full node was 160-odd GB in size. Now in April 2019, it is 210 GB in size. So this is the question I am going to pose to you. Would it be possible to use the data in the blockchain distributed database to identify patterns and statistical invariances to invest minimally with maximum possible profit? Can we forecast and build models to predict the prices of cryptocurrency in the future using AI and data science? The answer is a definite yes.
Practical Considerations
You may wonder if applying data science techniques and statistical analysis can actually produce information that can help in forecasting the future price of bitcoin. I came across a remarkable kernel on www.Kaggle.com (a website for data scientists to practice problems and compete with each other in competitions) by a user with the handle wayward artisan and the profile name Tania J. I thought it was worth sharing since this is a statistical analysis of the rise and the fall of the bitcoin bubble vividly illustrating how statistical methods helped this user to forecast the future price of bitcoin. The entire kernel is very large and interesting, please do visit it at the link given below. Just the start and the middle section of the kernel is given here because of space considerations and intellectual property considerations as well.
Your home for data science.
A Kaggle Kernel That Modelled the Bitcoin Bubble Burst Within Reasonable Error Limits
This following kernel uses cryptocurrency financial data scraped from www.coinmarketcap.com. It is a sobering example of how AI predictions actually predicted the collapse of the bitcoin bubble, prompting as many sellers to sell as they did. Coming across this kernel is one of the main motivations to write this article. I have omitted a lot of details, especially building the model and analyzing its accuracy. I just wanted to show that it was possible.
For more details, visit the kernel on Kaggle at the link: https://www.kaggle.com/taniaj/cryptocurrency-price-forecasting (Please visit this page, all aspiring data scientists. And pay attention to every concept discussed and used. Use Google and Wikipedia and you will learn a lot.)
A subset of the code is given below (the first section):
import pandas as pd
from pandas import DataFrame
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from datetime import datetime, timedelta
from statsmodels.tsa.arima_model import ARIMA
from statsmodels.tsa.statespace.sarimax import SARIMAX
from statsmodels.graphics.tsaplots import plot_acf, plot_pacf
from statsmodels.tsa.stattools import adfuller
from statsmodels.tsa.seasonal import seasonal_decompose
from scipy import stats
import statsmodels.api as sm
from itertools import product
from math import sqrt
from sklearn.metrics import mean_squared_error
import warnings
warnings.filterwarnings('ignore')
%matplotlib inline
colors = ["windows blue", "amber", "faded green", "dusty purple"]
sns.set(rc={"figure.figsize": (20,10), "axes.titlesize" : 18, "axes.labelsize" : 12,
"xtick.labelsize" : 14, "ytick.labelsize" : 14 })
<subsequent code not shown for brevity>
The dataset is available at the following link as a csv file in Microsoft Excel:
We focus on one of the middle sections with the first ARIMA model with SARIMAX (do look up Wikipedia and Google Search to learn about ARIMA and SARIMAX) which does the actual prediction at the time that the bitcoin bubble burst (only a subset of the code is shown). Visit the Kaggle kernel page on the link below this extract to get the entire code:
<data analysis and model analysis code section not shown here for brevity>
This code and the code earlier in the kernel (not shown for the sake of brevity) that built the model for accuracy gave the following predictions as output:
Bitcoin price forecasting at the time of the burst of the Bitcoin bubble
What do we learn? Surprisingly, the model captures the Bitcoin bubble burst with a remarkably accurate prediction (error levels ~ 10%)!
Conclusion
So, does AI and data science have anything to do with blockchain technology and cryptocurrency? The answer is a resounding, yes. Expect data science, statistical analysis, neural networks, and probability model distributions to play a heavy part when you want to forecast cryptocurrency prices.
For all the data science students out there, I am going to include one more screen from the same kernel on Kaggle (link):
The reason I want to show you this screen is that the terms and statistical lingo like kurtosis and heteroskedasticity are statistics concepts that you need to master in order to conduct forecasts like this, the main reason being to analyze the accuracy of the model you have constructed. The output window is given below:
So yes, blockchain technology and cryptocurrencies have a lot of overlap with applications. But also remember, data science can be applied to any field where finance is a factor.
Unless you’ve been living with your head under a rock for the last 4 years, you will definitely have heard of Bitcoin. You would also have heard about the technology behind Bitcoin, Blockchain. Now cryptocurrencies are banned in most cases in India and China, but the Americas and Europe still use cryptocurrencies extensively. And in my opinion, Asia stands to lose a lot if blockchain is not adopted extensively everywhere. Because make no mistake about it – blockchain technology will change the world as we know it. Forever.
Blockchain is the technology powering Bitcoin and other cryptocurrencies. To explain what blockchain is and what bitcoin is you can go through anyone of the articles below. Don’t worry these articles are carefully selected to be as interesting and fun to read as possible. (This also gives me space to add my own original ideas instead of copying or rewording existing articles – and I have plenty (of ideas)!
In fact, that last link is so amazingly simple visual and clear that I recommend everyone read it. Just so that we’re on the same page.
Exciting Applications
From Pixabay
Cut to the chase. A little confession here. I was asked to do this article nearly 16 days ago. Now I have some experience with blockchain before since having gone through it extensively as a research topic for my own blog. Then a remarkable idea hit me. An idea for a startup that could (in theory) become a multi-billion dollar enterprise. I spent a few days refining it, even going so far as to see if I could start this company with this area myself, until reality set in – I lacked the experience and the business skills.
No sooner had this realization struck me and the excitement cooled a little, another idea to improve blockchain struck me, and I promise to sketch out that idea as well. I am doing this for two reasons:
I am staunch support of the FOSS (free open source software movement and would like to be credited with the idea, and I am starting a free to use, open source project on GitHub – working on it, currently moving towards an alpha release as of now.
I believe in the power of technology to remove economic inequality. Now you may say that technology has evolved to the point that 4-5 monolithic companies dominate the entire world. But I believe that technology when used ethically has the potential to create more opportunities than it removes.
Blockchain has two major problems – energy consumption and resource consumption. But there are techniques that can alleviate both of these problems. We’ll deal with that as well in Part 2.
Finally, the vaunted hype about security for blockchain and cryptocurrencies is ridiculous when you think about it. For the sake of brevity, I will address the main security issues with blockchain in a separate article on Medium – (not here, since it has no relation to data science).
Application – A Personal Blockchain For Every Person On The Planet
In points (I assume you’ve gone through the graphical explanation of blockchain at least – if not you can review it here):
The trouble with end products of all types that are produced today is that there are so many intermediaries between the producer and the consumer that the producers receives a pittance compared to the end final price. It would be nice if we could track a product everywhere that it is used.
This is also applicable for books, music, articles, poems, pictures, any digital content of any sort. Currently Amazon and YouTube monopolize content distribution, the latter with a complete disregard for copyright and media ownership and payment. Suppose we had a tracking system that viewed every view of a video, and rewarded the original producer for it?
To emphasize the previous point, let us consider the case of Lindsey Stirling. Lindsey Stirling is a famous contemporary violinist who dances while playing. Her 118 video uploads have earned her 2,575,305,706 views, 2.5 billion approx, and her earnings from YouTube ads last month was 100K a month. Her net worth as on 10th April 2019 is 12 million USD (12,000,000).
But suppose Lindsey Stirling distributed her videos at a price of 1 USD every view. Her net worth would be 2.6 billion USD at the very least! She would be a multi-billionaire had this platform existed. It doesn’t – yet. And because it doesn’t exist she is 2.49 billion USD poorer!
Now everyone who knows blockchain technology will now realize this idea, the concept, and how blockchain can be used to overcome this problem – and its power. Disruptive power!
The Solution
The blockchain is a service that immutably assigns ownership.
The blockchain is also a database that stores every single transaction on a particular digitisable entity.
Finally, the Ethereum smart contract technology means that we can assign payments to go to every person on his own personal blockchain of all his digitisable goods.
This means we can build a world where producer pays a user-defined amount to every entity which created a particular digitisable product.
On this platform or website or marketplace, producers can adjust their prices and their payments and consumers can buy directly from them.
Everything can be tracked on the blockchain. Your own database of your own transactions can be used with smart contracts to pay the maximum possible fee to the most deserving person in the supply chain – fixed by each producer.
Hugely, Massively Disruptive
If you are interested or want to know more, you can leave a comment below with your email address. If you want to be a part of this new revolution and the new decentralised world – with all services provided free – please provide a comment below asking for my email ID with a statement of what and how you want to contribute to this endeavor. I promise to reply to every sincere query.
This is a fledgling project and a lot of work remains to be done. I will be writing articles and creating a team to work on this idea. Those of you who are interested please mail me at thomascherickal@gmail.com.
This will be an open source project and all services have to be offered free of cost. How do you go about making a profit from this? You don’t! The only way this can be fair to all players in countries like India is if it is specially designed to be applicable to anyone.
So this article gave a small glimpse into a world without intermediaries, corporations, money-making middlemen, and running purely on smart contracts. This is applicable to AI and data science since this technology will not reach anywhere significant without extensive use of AI and data science.
The more data that is available, the more analysis can be performed on it. And unless we have analysts who are running monitoring fraud detection systems fulltime on such a system, we might as well never build it – because blockchain data integrity cannot be hacked, but cryptocurrencies are hackable and have been hacked extensively since the beginning of Bitcoin.
For Part 2 of this series on Blockchain Applications of Data Science, you can go to the link below:
There are a huge number of ML algorithms out there. Trying to classify them leads to the distinction being made in types of the training procedure, applications, the latest advances, and some of the standard algorithms used by ML scientists in their daily work. There is a lot to cover, and we shall proceed as given in the following listing:
Statistical Algorithms
Classification
Regression
Clustering
Dimensionality Reduction
Ensemble Algorithms
Deep Learning
Reinforcement Learning
AutoML (Bonus)
1. Statistical Algorithms
Statistics is necessary for every machine learning expert. Hypothesis testing and confidence intervals are some of the many statistical concepts to know if you are a data scientist. Here, we consider here the phenomenon of overfitting. Basically, overfitting occurs when an ML model learns so many features of the training data set that the generalization capacity of the model on the test set takes a toss. The tradeoff between performance and overfitting is well illustrated by the following illustration:
Overfitting – from Wikipedia
Here, the black curve represents the performance of a classifier that has appropriately classified the dataset into two categories. Obviously, training the classifier was stopped at the right time in this instance. The green curve indicates what happens when we allow the training of the classifier to ‘overlearn the features’ in the training set. What happens is that we get an accuracy of 100%, but we lose out on performance on the test set because the test set will have a feature boundary that is usually similar but definitely not the same as the training set. This will result in a high error level when the classifier for the green curve is presented with new data. How can we prevent this?
Cross-Validation
Cross-Validation is the killer technique used to avoid overfitting. How does it work? A visual representation of the k-fold cross-validation process is given below:
From Quora
The entire dataset is split into equal subsets and the model is trained on all possible combinations of training and testing subsets that are possible as shown in the image above. Finally, the average of all the models is combined. The advantage of this is that this method eliminates sampling error, prevents overfitting, and accounts for bias. There are further variations of cross-validation like non-exhaustive cross-validation and nested k-fold cross validation (shown above). For more on cross-validation, visit the following link.
There are many more statistical algorithms that a data scientist has to know. Some examples include the chi-squared test, the Student’s t-test, how to calculate confidence intervals, how to interpret p-values, advanced probability theory, and many more. For more, please visit the excellent article given below:
Classification refers to the process of categorizing data input as a member of a target class. An example could be that we can classify customers into low-income, medium-income, and high-income depending upon their spending activity over a financial year. This knowledge can help us tailor the ads shown to them accurately when they come online and maximises the chance of a conversion or a sale. There are various types of classification like binary classification, multi-class classification, and various other variants. It is perhaps the most well known and most common of all data science algorithm categories. The algorithms that can be used for classification include:
Logistic Regression
Support Vector Machines
Linear Discriminant Analysis
K-Nearest Neighbours
Decision Trees
Random Forests
and many more. A short illustration of a binary classification visualization is given below:
From openclassroom.stanford.edu
For more information on classification algorithms, refer to the following excellent links:
Regression is similar to classification, and many algorithms used are similar (e.g. random forests). The difference is that while classification categorizes a data point, regression predicts a continuous real-number value. So classification works with classes while regression works with real numbers. And yes – many algorithms can be used for both classification and regression. Hence the presence of logistic regression in both lists. Some of the common algorithms used for regression are
Linear Regression
Support Vector Regression
Logistic Regression
Ridge Regression
Partial Least-Squares Regression
Non-Linear Regression
For more on regression, I suggest that you visit the following link for an excellent article:
Both articles have a remarkably clear discussion of the statistical theory that you need to know to understand regression and apply it to non-linear problems. They also have source code in Python and R that you can use.
4. Clustering
Clustering is an unsupervised learning algorithm category that divides the data set into groups depending upon common characteristics or common properties. A good example would be grouping the data set instances into categories automatically, the process being used would be any of several algorithms that we shall soon list. For this reason, clustering is sometimes known as automatic classification. It is also a critical part of exploratory data analysis (EDA). Some of the algorithms commonly used for clustering are:
Hierarchical Clustering – Agglomerative
Hierarchical Clustering – Divisive
K-Means Clustering
K-Nearest Neighbours Clustering
EM (Expectation Maximization) Clustering
Principal Components Analysis Clustering (PCA)
An example of a common clustering problem visualization is given below:
From Wikipedia
The above visualization clearly contains three clusters.
Another excellent article on clustering refer the link
Dimensionality Reduction is an extremely important tool that should be completely clear and lucid for any serious data scientist. Dimensionality Reduction is also referred to as feature selection or feature extraction. This means that the principal variables of the data set that contains the highest covariance with the output data are extracted and the features/variables that are not important are ignored. It is an essential part of EDA (Exploratory Data Analysis) and is nearly always used in every moderately or highly difficult problem. The advantages of dimensionality reduction are (from Wikipedia):
It reduces the time and storage space required.
Removal of multi-collinearity improves the interpretation of the parameters of the machine learning model.
It becomes easier to visualize the data when reduced to very low dimensions such as 2D or 3D.
It avoids the curse of dimensionality.
The most commonly used algorithm for dimensionality reduction is Principal Components Analysis or PCA. While this is a linear model, it can be converted to a non-linear model through a kernel trick similar to that used in a Support Vector Machine, in which case the technique is known as Kernel PCA. Thus, the algorithms commonly used are:
Ensembling means combining multiple ML learners together into one pipeline so that the combination of all the weak learners makes an ML application with higher accuracy than each learner taken separately. Intuitively, this makes sense, since the disadvantages of using one model would be offset by combining it with another model that does not suffer from this disadvantage. There are various algorithms used in ensembling machine learning models. The three common techniques usually employed in practice are:
Simple/Weighted Average/Voting: Simplest one, just takes the vote of models in Classification and average in Regression.
Bagging: We train models (same algorithm) in parallel for random sub-samples of data-set with replacement. Eventually, take an average/vote of obtained results.
Boosting: In this models are trained sequentially, where (n)th model uses the output of (n-1)th model and works on the limitation of the previous model, the process stops when result stops improving.
Stacking: We combine two or more than two models using another machine learning algorithm.
(from Amardeep Chauhan on Medium.com)
In all four cases, the combination of the different models ends up having the better performance that one single learner. One particular ensembling technique that has done extremely well on data science competitions on Kaggle is the GBRT model or the Gradient Boosted Regression Tree model.
We include the source code from the scikit-learn module for Gradient Boosted Regression Trees since this is one of the most popular ML models which can be used in competitions like Kaggle, HackerRank, and TopCoder.
GradientBoostingClassifier supports both binary and multi-class classification. The following example shows how to fit a gradient boosting classifier with 100 decision stumps as weak learners:
GradientBoostingRegressor supports a number of different loss functions for regression which can be specified via the argument loss; the default loss function for regression is least squares ('ls').
import numpy as np
from sklearn.metrics import mean_squared_error
from sklearn.datasets import make_friedman1
from sklearn.ensemble import GradientBoostingRegressor
X, y = make_friedman1(n_samples=1200, random_state=0, noise=1.0)
X_train, X_test = X[:200], X[200:]
y_train, y_test = y[:200], y[200:]
est = GradientBoostingRegressor(n_estimators=100, learning_rate=0.1,
max_depth=1, random_state=0, loss='ls').fit(X_train, y_train)
mean_squared_error(y_test, est.predict(X_test))
You can also refer to the following article which discusses Random Forests, which is a (rather basic) ensembling method.
In the last decade, there has been a renaissance of sorts within the Machine Learning community worldwide. Since 2002, neural networks research had struck a dead end as the networks of layers would get stuck in local minima in the non-linear hyperspace of the energy landscape of a three layer network. Many thought that neural networks had outlived their usefulness. However, starting with Geoffrey Hinton in 2006, researchers found that adding multiple layers of neurons to a neural network created an energy landscape of such high dimensionality that local minima were statistically shown to be extremely unlikely to occur in practice. Today, in 2019, more than a decade of innovation later, this method of adding addition hidden layers of neurons to a neural network is the classical practice of the field known as deep learning.
Deep Learning has truly taken the computing world by storm and has been applied to nearly every field of computation, with great success. Now with advances in Computer Vision, Image Processing, Reinforcement Learning, and Evolutionary Computation, we have marvellous feats of technology like self-driving cars and self-learning expert systems that perform enormously complex tasks like playing the game of Go (not to be confused with the Go programming language). The main reason these feats are possible is the success of deep learning and reinforcement learning (more on the latter given in the next section below). Some of the important algorithms and applications that data scientists have to be aware of in deep learning are:
Long Short term Memories (LSTMs) for Natural Language Processing
Recurrent Neural Networks (RNNs) for Speech Recognition
Convolutional Neural Networks (CNNs) for Image Processing
Deep Neural Networks (DNNs) for Image Recognition and Classification
Hybrid Architectures for Recommender Systems
Autoencoders (ANNs) for Bioinformatics, Wearables, and Healthcare
Deep Learning Networks typically have millions of neurons and hundreds of millions of connections between neurons. Training such networks is such a computationally intensive task that now companies are turning to the 1) Cloud Computing Systems and 2) Graphical Processing Unit (GPU) Parallel High-Performance Processing Systems for their computational needs. It is now common to find hundreds of GPUs operating in parallel to train ridiculously high dimensional neural networks for amazing applications like dreaming during sleep and computer artistry and artistic creativity pleasing to our aesthetic senses.
Artistic Image Created By A Deep Learning Network. From blog.kadenze.com.
For more on Deep Learning, please visit the following links:
In the recent past and the last three years in particular, reinforcement learning has become remarkably famous for a number of achievements in cognition that were earlier thought to be limited to humans. Basically put, reinforcement learning deals with the ability of a computer to teach itself. We have the idea of a reward vs. penalty approach. The computer is given a scenario and ‘rewarded’ with points for correct behaviour and ‘penalties’ are imposed for wrong behaviour. The computer is provided with a problem formulated as a Markov Decision Process, or MDP. Some basic types of Reinforcement Learning algorithms to be aware of are (some extracts from Wikipedia):
1.Q-Learning
Q-Learning is a model-free reinforcement learning algorithm. The goal of Q-learning is to learn a policy, which tells an agent what action to take under what circumstances. It does not require a model (hence the connotation “model-free”) of the environment, and it can handle problems with stochastic transitions and rewards, without requiring adaptations. For any finite Markov decision process (FMDP), Q-learning finds a policy that is optimal in the sense that it maximizes the expected value of the total reward over any and all successive steps, starting from the current state. Q-learning can identify an optimal action-selection policy for any given FMDP, given infinite exploration time and a partly-random policy. “Q” names the function that returns the reward used to provide the reinforcement and can be said to stand for the “quality” of an action taken in a given state.
2.SARSA
State–action–reward–state–action (SARSA) is an algorithm for learning a Markov decision process policy. This name simply reflects the fact that the main function for updating the Q-value depends on the current state of the agent “S1“, the action the agent chooses “A1“, the reward “R” the agent gets for choosing this action, the state “S2” that the agent enters after taking that action, and finally the next action “A2” the agent choose in its new state. The acronym for the quintuple (st, at, rt, st+1, at+1) is SARSA.
3.Deep Reinforcement Learning
This approach extends reinforcement learning by using a deep neural network and without explicitly designing the state space. The work on learning ATARI games by Google DeepMind increased attention to deep reinforcement learning or end-to-end reinforcement learning. Remarkably, the computer agent DeepMind has achieved levels of skill higher than humans at playing computer games. Even a complex game like DOTA 2 was won by a deep reinforcement learning network based upon DeepMind and OpenAI Gym environments that beat human players 3-2 in a tournament of best of five matches.
For more information, go through the following links:
If reinforcement learning was cutting edge data science, AutoML is bleeding edge data science. AutoML (Automated Machine Learning) is a remarkable project that is open source and available on GitHub at the following link that, remarkably, uses an algorithm and a data analysis approach to construct an end-to-end data science project that does data-preprocessing, algorithm selection,hyperparameter tuning, cross-validation and algorithm optimization to completely automate the ML process into the hands of a computer. Amazingly, what this means is that now computers can handle the ML expertise that was earlier in the hands of a few limited ML practitioners and AI experts.
AutoML has found its way into Google TensorFlow through AutoKeras, Microsoft CNTK, and Google Cloud Platform, Microsoft Azure, and Amazon Web Services (AWS). Currently it is a premiere paid model for even a moderately sized dataset and is free only for tiny datasets. However, one entire process might take one to two or more days to execute completely. But at least, now the computer AI industry has come full circle. We now have computers so complex that they are taking the machine learning process out of the hands of the humans and creating models that are significantly more accurate and faster than the ones created by human beings!
The basic algorithm used by AutoML is Network Architecture Search and its variants, given below:
Network Architecture Search (NAS)
PNAS (Progressive NAS)
ENAS (Efficient NAS)
The functioning of AutoML is given by the following diagram:
If you’ve stayed with me till now, congratulations; you have learnt a lot of information and cutting edge technology that you must read up on, much, much more. You could start with the links in this article, and of course, Google is your best friend as a Machine Learning Practitioner. Enjoy machine learning!
2019 looks to be the year of using smarter technology in a smarter way. Three key trends — artificial intelligence systems becoming a serious component in enterprise tools, custom hardware breaking out for special use-cases, and a rethink on data science and its utility — will all combine into a common theme.
In recent years, we’ve seen all manner of jaw-dropping technology, but the emphasis has been very much on what these gadgets and systems can do and how they do it, with much less attention paid to why.
In this blog, we will explore different areas in data science and figure out our expectations in 2019 in them. Areas include machine learning, AR/VR systems, edge computing etc. Let us go through them one by one
Machine Learning/Deep Learning
Businesses are using machine learning to improve all sorts of outcomes, from optimizing operational workflows and increasing customer satisfaction to discovering to a new competitive differentiator. But, now all the hype around AI is settling. Machine learning is not a cool term anymore. Furthermore, organisations are looking for more ways of identifying more options in the form of agent modelling. Apart from this, more adoption of these algorithms looks very feasible now. Adoption will be seen in new and old industries
Healthcare companies are already big users of AI, and this trend will continue. According to Accenture, the AI healthcare market might hit $6.6 billion by 2021, and clinical health AI applications can create $150 billion in annual savings for the U.S. healthcare economy by 2026.
In retail, global spending on AI will grow to $7.3 billion a year by 2022, up from $2 billion in 2018, according to Juniper Research. This is because companies will invest heavily in AI tools that will help them differentiate and improve the services they offer customers.
In cybersecurity, the adoption of AI brings a boom in startups that are able to raised$3.65 billion in equity funding in the last five years. Cyber AI can help security experts sort through millions of incidents to identify aberrations, risks, and signals of future threats.
And there is even an opportunity brewing in industries facing labour shortages, such as transportation. At the end of 2017, there was a shortage of 51,000 truck drivers (up from a shortage of 36,000 the previous year). And the ATA reports that the trucking industry will need to hire 900,000 more drivers in the next 10 years to keep up with demand. AI-driven autonomous vehicles could help relieve the need for more drivers in the future.
Programming Language
The practice of data science requires the use of analytics tools, technologies and programming languages to help data professionals extract insights and value from data. A recent survey of nearly 24,000 data professionals by Kaggle suggests that Python, SQL and R are the most popular programming languages. The most popular, by far, was Python (83%). Additionally, 3 out of 4 data professionals recommended that aspiring data scientists learn Python first.
Survey results show that 3 out of 4 data professionals would recommend Python as the programming language aspiring data scientists to learn first. The remaining programming languages are recommended at a significantly lower rate (R recommended by 12% of respondents; SQL by 5% of respondents. Anyhow, Python will also boom more in 2019. But, R community too have come up with a lot of recent advancements. With new packages and improvements, R is expected to come closer to python in terms of usage.
Blockchain and Big Data
In recent years, the blockchain is at the heart of computer technologies. It is a cryptographically secure distributed database technology for storing and transmitting information. The main advantage of the blockchain is that it is decentralized. In fact, no one controls the data entering or their integrity. However, these checks run through various computers on the network. These different machines hold the same information. In fact, faulty data on one computer cannot enter the chain because it will not match the equivalent data held by the other machines. To put it simply, as long as the network exists, the information remains in the same state.
Big Data analytics will be essential for tracking transactions and enabling businesses that use the Blockchain to make better decisions. That’s why new Data Intelligence services are emerging to help financial institutions and governments and other businesses discover who they interact with within the Blockchain and discover hidden patterns.
Augmented-Reality/Virtual Reality
The broader the canvas of visualization is, the better the understanding is. That’s exactly what happens when one visualizes big data through the Augmented Reality (AR) and Virtual Reality (VR). A combination of AR and VR could open a world of possibilities to better utilize the data at hand. VR and AR can practically improve the way we perceive data and could actually be the solution to make use of the large unused data.
By presenting the data in the form of 3D, the user will be able to decipher the major takeaways from the data better and faster with easier understanding. Many recent types of research show that the VR and AR has a high sensory impact which promotes faster learning and understanding.
This immersive way of representation of the data enables the analysts to handle the big data more efficiently. It makes the analysis and interpretation more of an experience and realisation that the traditional analysis. Instead of the user seeing numbers and figures, the person will be able to see beyond it and into the facts, happenings and reasons which could revolutionize the businesses.
Edge Computing
Computing infrastructure is an ever-changing landscape of technology advancements. Current changes affect the way companies deploy smart manufacturing systems to make the most of advancements.
The rise of edge computing capabilities coupled with traditional industrial control system (ICS) architectures provides increasing levels of flexibility. In addition, time-synchronized applications and analytics augment the need for larger Big Data operations in the cloud. This is regardless of cloud premise.
Edge is still in early stage adoption. But, one thing is clear that edge devices are subject to large-scale investments from cloud suppliers to offload bandwidth. Also, there are latency issues due to an explosion of the IoT data in both industrial and commercial applications.
Edge soon will likely increase in adoption where users have questions about the cloud’s specific use case. Cloud-level interfaces and apps will migrate to the edge. Industrial application hosting and analytics will become common at the edge. This will happen using virtual servers and simplified operational technology-friendly hardware and software.
The Rise of Semi-Automated Tools for Data Science
There has been a rise of self-service BI tools such as Tableau, Qlik Sense, Power BI, and Domo. Furthermore, now managers can obtain current business information in graphical form on demand. Although, IT may need to set up a certain amount of setup at the outset. Also, when adding a data source, most of the data cleaning work and analysis can be done by analysts. The analyses can update automatically from the latest data any time they are opened.
Managers can then interact with the analyses graphically to identify issues that need to be addressed. In a BI-generated dashboard or “story” about sales numbers, that might mean drilling down to find underperforming stores, salespeople, and products, or discovering trends in year-over-year same-store comparisons. These discoveries might in turn guide decisions about future stocking levels, product sales and promotions. Also, they may determine the building of additional stores in under-served areas.
Upgrade in Job Roles
In recent times, there have been a lot of advancements in the data science industry. With these advancements, different businesses are in better shape to extract much more value out of their data. With an increase in expectation, there is a shift in the roles of both data scientists and business analysts now. The data scientists should move from statistical focus phase to more of a research phase. But the business analysts are now filling in the gap left by data scientists and are taking their roles up.
We can see it as an upgrade in both the job roles. Business analysts now hold the business angle firm but are also handling the statistical and technical part of the things too. Business analysts are now more into predictive analytics. They are at a stage now where they can use off-the-shelf algorithms for predictions in their business domains. BA’s are not only for reporting and business mindset but now are more into the prescriptive analytics too. They are handling the role of model building, data warehousing and statistical analysing.
Summary
How this question is answered will be fascinating to watch. It could be that the data science field has to completely overhaul what it can offer, overcoming seeming off-limit barriers. Alternatively, it could be that businesses discover their expectations can’t be met and have to adjust to this reality in a productive manner rather than get bogged down in frustration.
In conclusion, 2019 promises to be a year where smart systems make further inroads into our personal and professional lives. More importantly, I expect our professional lives to get more sophisticated with a variety of agents and systems helping us get more of out of our time in the office!
Follow this link, if you are looking to learn more about data science online!
Technology has become the embedded component of applications and the defacto driver for growth in industries. With the advent of AI, new milestones are being achieved each day. We are moving towards an era of more and more integration, making it an indispensable mediator between systems and humans. The rapid strides taken by the mobile industry seems like an overwhelming convergence of multiple worlds. The innate ability of such systems to improve itself, strengthened by data analytics, IoT and AI have opened new frontiers. To reap the unbound merits of AI, software application vendors are integrating it into software applications.
In this blog, we will understand what exactly are these intelligent apps. What all does it require to make an intelligent app? Also, we will look into the real world applications of these intelligent apps.
What are Intelligent Applications?
So what exactly are intelligent apps? These are apps that not only know how to support key user decisions but also learn from user interactions. These apps aim to become even more relevant and valuable to these users.
In other words, intelligent apps are those that also learn and adapt and can even act on their own. Much like all of us, these apps learn and change behaviour. We are already seeing this at work. Have you noticed how e-commerce websites show you the right recommendations at the right time?
Intelligent apps are becoming a thing thanks to the strides being made in Artificial Intelligence (AI) and Machine Learning. Machine learning gives systems the ability to learn and improve from experience without being specifically programmed. There is an increase in the popularity of conversational systems and the growth of the Internet of Things. Therefore, we are seeing machine learning applied to more things in our everyday life.
Using AI algorithms, intelligent apps can study users’ behaviour and choices. Furthermore, it can sort through this data to use the relevant information to predict your needs and act on your behalf. For example, Smart Reply enables you to quickly respond to emails by giving you auto-generated replies. Productivity apps like Microsoft Office 365 and Google’s G Suite also use AI. Chatbots such as Meziuse machine learning to study user’s behaviour and provide them with choices they would like.
Features of Intelligent Applications
1. Data-driven
Intelligent apps combine and process multiple data sources — such as IoT sensors, beacons or user interactions — and turn an enormous quantity of numbers into valuable insights.
2. Contextual and relevant
Intelligent apps make much smarter use of a device’s features to proactively deliver highly relevant information and suggestions. Users will no longer have to go to their apps. Instead, the apps will come to them.
3. Continuously adapting
Through machine learning, intelligent apps continuously adapt and improve their output.
4. Action-oriented
By anticipating user behaviours with predictive analytics, smart applications deliver personalized and actionable suggestions.
5. Omnichannel
Progressive web applications are increasingly blurring the lines between native apps and mobile web applications.
Applications
1. Health Care Benefits
We are exploring AI/ML technology for health care. It can help doctors with diagnoses and tell when patients are deteriorating so medical intervention can occur sooner before the patient needs hospitalization. It’s a win-win for the healthcare industry, saving costs for both the hospitals and patients. The precision of machine learning can also detect diseases such as cancer sooner, thus saving lives.
2. Intelligent Conversational Interfaces
We are using machine learning and AI to build intelligent conversational chatbots and voice skills. These AI-driven conversational interfaces are answering questions from frequently asked questions and answers, helping users with concierge services in hotels, and to provide information about products for shopping. Advancements in the deep neural network or deep learning are making many of these AI and ML applications possible
3. Market Prediction
We are using AI in a number of traditional places like personalization, intuitive workflows, enhanced searching and product recommendations. More recently, we started baking AI into our go-to-market operations to be first to market by predicting the future. Or should I say, by “trying” to predict the future?
4. Customer Lifetime Value Modeling
Customer lifetime value models are among the most important for eCommerce business to employ. That’s because they can be used to identify, understand, and retain your company’s most valuable customers, whether that means the biggest spenders, the most loyal advocates of your brand, or both. These models predict the future revenue that an individual customer will bring to your business in a given period. With this information, you can focus your marketing efforts to encourage these customers to interact with your brand more often and even target your acquisition spend to attract new customers that are similar to your existing MVPs.
5. Churn Modeling
Customer churn modelling can help you identify which of your customers are likely to stop engaging with your business and why. The results of a churn model can range from churn risk scores for individual customers to drivers of churn ranked by importance. These outputs are essential components of an algorithmic retention strategy because they help optimize discount offers, email campaigns, or other targeted marketing initiatives that keep your high-value customer’s buying.
6. Dynamic Pricing
Dynamic pricing, also known as demand pricing, is the practice of flexible pricing items based on factors like the level of interest of the target customer, demand at the time of purchase, or whether the customer has engaged with a marketing campaign. This requires a lot of data about how different customers’ willingness to pay for a good or service changes across a variety of situations, but companies like airlines and ride-share services have successfully implemented dynamic price optimization strategies to maximize revenue.
7. Customer Segmentation
Data scientists do not rely on intuition to separate customers into groups. They use clustering and classification algorithms to group customers into personas based on specific variations among them. These personas account for customer differences across multiple dimensions such as demographics, browsing behaviour, and affinity. Connecting these traits to patterns of purchasing behaviour allows data-savvy companies to roll out highly personalized marketing campaigns. Additionally, these campaigns are more effective at boosting sales than generalized campaigns.
8. Image Classification
Image classification uses machine learning algorithms to assign a label from a fixed set of categories to an image that’s inputted. It has a wide range of business applications including modelling 3D construction plans based on 2D designs, social media photo tagging, informing medical diagnoses, and more. Deep learning methods such as neural networks are often used for image classification because they can most effectively identify relevant features of an image in the presence of potential complications like the variation in the point of view, illumination, scale, or volume of clutter in the image.
9. Recommendation Engines
Recommendation engines are another major way machine learning proves its business value. In fact, Netflix values the recommendation engine powering its content suggestions at $1 billion per year and Amazon says its system drives a 20–35% lift in sales annually. That’s because recommendation engines sift through large quantities of data to predict how likely any given customer is to purchase an item or enjoy a piece of content and then suggest those things to the user. The result is a customer experience that encourages better engagement and reduces churn.
Examples
1. Email Filters in Gmail
Google uses AI to ensure that nearly all of the email landing in your inbox is authentic. Their filters attempt to sort emails into the following categories like primary, social, promotions, updates, forums and spam. The program helps your emails get organized so you can find your way to important communications quicker.
2. LinkedIn
AI is used to help match candidates to jobs with the hopes of creating better employee-employer matches.
On its talent blog, LinkedIn explains that they use “deeper insights into the behaviour of applicants on LinkedIn” in order to “predict not just who would apply to your job, but who would get hired…”
3. Google Predictive Searches
When you begin typing a search term and Google makes recommendations for you to choose from, that’s AI in action. Predictive searches are based on data that Google collects about you, such as your location, age, and other personal details. Using AI, the search engine attempts to guess what you might be trying to find.
4. Tesla Smart Cars
Talking about the AI, there is no better and more prominent display of this technology than what smart car and drone manufacturers are doing with it. Just a few years back, using a fully automatic car was a dream, however, now companies like Tesla have made so much progress that we already have a fleet of semi-automatic cars on the road.
5. Online Ads Network(Facebook/Microsoft/Google)
One of the biggest users of artificial intelligence is the online ad industry which uses AI to not only track user statistics but also serve us ads based on those statistics. Without AI, the online ad industry will just fail as it would show random ads to users with no connection to their preferences what so ever. AI has become so successful in determining our interests and serving us ads that the global digital ad industry has crossed 250 billion US dollars with the industry projected to cross the 300 billion mark in 2019. So next time when you are going online and seeing ads or product recommendation, know that AI is impacting your life.
6. Amazon Product Recommendations
Amazon and other online retailers use AI to gather information about your preferences and buying habits. Then, they personalize your shopping experience by suggesting new products tailored to your habits.
When you search for an item such as “Bose headsets,” the search engine also shows related items that other people have purchased when searching for the same product.
Current trends and explorations
Intelligent things are poised to be one of the important trends that have the potential for ‘disruption’ and large-scale impact across industries. According to Gartner, the future will see the utilization of AI by almost all apps and services, making these apps discreet yet useful and intelligent mediators between systems and humans. AI will be incorporated into various systems and apps in some way and is poised to become the key enabler across a variety of services and software systems. As mentioned at the Google conference, very fast, we are moving from mobile-first to AI-first world.
It won’t be an exaggeration to say that all the new applications built in the coming years will be intelligent apps. These apps use machine learning and historical as well as real-time data to make smart decisions and deliver a highly personalized experience to the users. These apps combine predictive and prescriptive analytics, customer data, product insights, and operational vision with contemporary user-focused design and application development tools to create a highly impactful experience for users.
The intelligent apps undoubtedly have the potential to change the face of work and structure at companies in the coming years. According to Gartner’s prediction, companies will increasingly use and develop intelligent apps and utilize analytics and big data to enhance their business processes and offer top class customer experiences.
Summary
As companies are charting their digital transformation initiatives, they need to add intelligent apps to their blueprint. The development of the right intelligent apps needs to consider the new growth areas, internal and external data sources, real-time data acquisition, processing, and analysis and putting the right technology to use.
Intelligent apps are undoubtedly paving the way for speedier business decisions, better business results, greater efficiency of the workforce, and long-term gains for all — they just need to be utilized right. Companies which are diving in intelligent apps now will have a considerable competitive advantage in the near future.
Follow this link, if you are looking to learn more about data science online!
Never thought that online trading could be so helpful because of so many scammers online until I met Miss Judith... Philpot who changed my life and that of my family. I invested $1000 and got $7,000 Within a week. she is an expert and also proven to be trustworthy and reliable. Contact her via: Whatsapp: +17327126738 Email:judithphilpot220@gmail.comread more
A very big thank you to you all sharing her good work as an expert in crypto and forex trade option. Thanks for... everything you have done for me, I trusted her and she delivered as promised. Investing $500 and got a profit of $5,500 in 7 working days, with her great skill in mining and trading in my wallet.
judith Philpot company line:... WhatsApp:+17327126738 Email:Judithphilpot220@gmail.comread more
Faculty knowledge is good but they didn't cover most of the topics which was mentioned in curriculum during online... session. Instead they provided recorded session for those.read more
Dimensionless is great place for you to begin exploring Data science under the guidance of experts. Both Himanshu and... Kushagra sir are excellent teachers as well as mentors,always available to help students and so are the HR and the faulty.Apart from the class timings as well, they have always made time to help and coach with any queries.I thank Dimensionless for helping me get a good starting point in Data science.read more
My experience with the data science course at Dimensionless has been extremely positive. The course was effectively... structured . The instructors were passionate and attentive to all students at every live sessions. I could balance the missed live sessions with recorded ones. I have greatly enjoyed the class and would highly recommend it to my friends and peers.
Special thanks to the entire team for all the personal attention they provide to query of each and every student.read more
It has been a great experience with Dimensionless . Especially from the support team , once you get enrolled , you... don't need to worry about anything , they keep updating each and everything. Teaching staffs are very supportive , even you don't know any thing you can ask without any hesitation and they are always ready to guide . Definitely it is a very good place to boost careerread more
The training experience has been really good! Specially the support after training!! HR team is really good. They keep... you posted on all the openings regularly since the time you join the course!! Overall a good experience!!read more
Dimensionless is the place where you can become a hero from zero in Data Science Field. I really would recommend to all... my fellow mates. The timings are proper, the teaching is awsome,the teachers are well my mentors now. All inclusive I would say that Kush Sir, Himanshu sir and Pranali Mam are the real backbones of Data Science Course who could teach you so well that even a person from non- Math background can learn it. The course material is the bonus of this course and also you will be getting the recordings of every session. I learnt a lot about data science and Now I find it easy because of these wonderful faculty who taught me. Also you will get the good placement assistance as well as resume bulding guidance from Venu Mam. I am glad that I joined dimensionless and also looking forward to start my journey in data science field. I want to thank Dimensionless because of their hard work and Presence it made it easy for me to restart my career. Thank you so much to all the Teachers in Dimensionless !read more
Dimensionless has great teaching staff they not only cover each and every topic but makes sure that every student gets... the topic crystal clear. They never hesitate to repeat same topic and if someone is still confused on it then special doubt clearing sessions are organised. HR is constantly busy sending us new openings in multiple companies from fresher to Experienced. I would really thank all the dimensionless team for showing such support and consistency in every thing.read more
I had great learning experience with Dimensionless. I am suggesting Dimensionless because of its great mentors... specially Kushagra and Himanshu. they don't move to next topic without clearing the concept.read more
My experience with Dimensionless has been very good. All the topics are very well taught and in-depth concepts are... covered. The best thing is that you can resolve your doubts quickly as its a live one on one teaching. The trainers are very friendly and make sure everyone's doubts are cleared. In fact, they have always happily helped me with my issues even though my course is completed.read more
I would highly recommend dimensionless as course design & coaches start from basics and provide you with a real-life... case study. Most important is efforts by all trainers to resolve every doubts and support helps make difficult topics easy..read more
Dimensionless is great platform to kick start your Data Science Studies. Even if you are not having programming skills... you will able to learn all the required skills in this class.All the faculties are well experienced which helped me alot. I would like to thanks Himanshu, Pranali , Kush for your great support. Thanks to Venu as well for sharing videos on timely basis...😊
I highly recommend dimensionless for data science training and I have also been completed my training in data science... with dimensionless. Dimensionless trainer have very good, highly skilled and excellent approach. I will convey all the best for their good work. Regards Avneetread more
After a thinking a lot finally I joined here in Dimensionless for DataScience course. The instructors are experienced &... friendly in nature. They listen patiently & care for each & every students's doubts & clarify those with day-to-day life examples. The course contents are good & the presentation skills are commendable. From a student's perspective they do not leave any concept untouched. The step by step approach of presenting is making a difficult concept easier. Both Himanshu & Kush are masters of presenting tough concepts as easy as possible. I would like to thank all instructors: Himanshu, Kush & Pranali.read more
When I start thinking about to learn Data Science, I was trying to find a course which can me a solid understanding of... Statistics and the Math behind ML algorithms. Then I have come across Dimensionless, I had a demo and went through all my Q&A, course curriculum and it has given me enough confidence to get started. I have been taught statistics by Kush and ML from Himanshu, I can confidently say the kind of stuff they deliver is In depth and with ease of understanding!read more
If you love playing with data & looking for a career change in Data science field ,then Dimensionless is the best... platform . It was a wonderful learning experience at dimensionless. The course contents are very well structured which covers from very basics to hardcore . Sessions are very interactive & every doubts were taken care of. Both the instructors Himanshu & kushagra are highly skilled, experienced,very patient & tries to explain the underlying concept in depth with n number of examples. Solving a number of case studies from different domains provides hands-on experience & will boost your confidence. Last but not the least HR staff (Venu) is very supportive & also helps in building your CV according to prior experience and industry requirements. I would love to be back here whenever i need any training in Data science further.read more
It was great learning experience with statistical machine learning using R and python. I had taken courses from... Coursera in past but attention to details on each concept along with hands on during live meeting no one can beat the dimensionless team.read more
I would say power packed content on Data Science through R and Python. If you aspire to indulge in these newer... technologies, you have come at right place. The faculties have real life industry experience, IIT grads, uses new technologies to give you classroom like experience. The whole team is highly motivated and they go extra mile to make your journey easier. I’m glad that I was introduced to this team one of my friends and I further highly recommend to all the aspiring Data Scientists.read more
It was an awesome experience while learning data science and machine learning concepts from dimensionless. The course... contents are very good and covers all the requirements for a data science course. Both the trainers Himanshu and Kushagra are excellent and pays personal attention to everyone in the session. thanks alot !!read more
Had a great experience with dimensionless.!! I attended the Data science with R course, and to my finding this... course is very well structured and covers all concepts and theories that form the base to step into a data science career. Infact better than most of the MOOCs. Excellent and dedicated faculties to guide you through the course and answer all your queries, and providing individual attention as much as possible.(which is really good). Also weekly assignments and its discussion helps a lot in understanding the concepts. Overall a great place to seek guidance and embark your journey towards data science.read more
Excellent study material and tutorials. The tutors knowledge of subjects are exceptional. The most effective part... of curriculum was impressive teaching style especially that of Himanshu. I would like to extend my thanks to Venu, who is very responsible in her jobread more
It was a very good experience learning Data Science with Dimensionless. The classes were very interactive and every... query/doubts of students were taken care of. Course structure had been framed in a very structured manner. Both the trainers possess in-depth knowledge of data science dimain with excellent teaching skills. The case studies given are from different domains so that we get all round exposure to use analytics in various fields. One of the best thing was other support(HR) staff available 24/7 to listen and help.I recommend data Science course from Dimensionless.read more
I was a part of 'Data Science using R' course. Overall experience was great and concepts of Machine Learning with R... were covered beautifully. The style of teaching of Himanshu and Kush was quite good and all topics were generally explained by giving some real world examples. The assignments and case studies were challenging and will give you exposure to the type of projects that Analytics companies actually work upon. Overall experience has been great and I would like to thank the entire Dimensionless team for helping me throughout this course. Best wishes for the future.read more
It was a great experience leaning data Science with Dimensionless .Online and interactive classes makes it easy to... learn inspite of busy schedule. Faculty were truly remarkable and support services to adhere queries and concerns were also very quick. Himanshu and Kush have tremendous knowledge of data science and have excellent teaching skills and are problem solving..Help in interviews preparations and Resume building...Overall a great learning platform. HR is excellent and very interactive. Everytime available over phone call, whatsapp, mails... Shares lots of job opportunities on the daily bases... guidance on resume building, interviews, jobs, companies!!!! They are just excellent!!!!! I would recommend everyone to learn Data science from Dimensionless only 😊read more
Being a part of IT industry for nearly 10 years, I have come across many trainings, organized internally or externally,... but I never had the trainers like Dimensionless has provided. Their pure dedication and diligence really hard to find. The kind of knowledge they possess is imperative. Sometimes trainers do have knowledge but they lack in explaining them. Dimensionless Trainers can give you ‘N’ number of examples to explain each and every small topic, which shows their amazing teaching skills and In-Depth knowledge of the subject. Himanshu and Kush provides you the personal touch whenever you need. They always listen to your problems and try to resolve them devotionally.
I am glad to be a part of Dimensionless and will always come back whenever I need any specific training in Data Science. I recommend this to everyone who is looking for Data Science career as an alternative.
All the best guys, wish you all the success!!read more
