AI News, Data Science, Machine Learning, and Statistics: what is in a name?

Data Science, Machine Learning, and Statistics: what is in a name?

We (the authors on this blog) label many of our articles as being about data science because we want to emphasize that the various techniques we write about are only meaningful when considered parts of a larger end to end process.

The important components are learning the true business needs (often by extensive partnership with customers), enabling the collection of data, managing data, applying modeling techniques and applying statistics criticisms.

The pre-existing term I have found that is closest to describing this whole project system is data science, so that is the term I use.

group in the mid 1990’s (this group emphasized simulation based modeling of small molecules and their biological interactions, the naming was an attempt to emphasize computation over computers).

I think there are enough substantial differences in approach between traditional statistics, machine learning, data mining, predictive analytics, and data science to justify at least this much nomenclature.

There are many other potential data problems statistics describes well (like Simpson’s paradox), but statistics is fairly unique in the information sciences in emphasizing the risks of trying to reason from small datasets.

It is only recently that minimally curated big datasets became perceived as being inherently valuable (the earlier attitude being closer to GIGO).

Often a big dataset (such as logs of all clicks seen on a search engine) is useful largely because they are a good proxy for a smaller dataset that is too expensive to actually produce (such as interviewing a good cross section of search engine users as to their actual intent).

Good machine learning papers use good optimization techniques and bad machine learning papers (most of them in fact) use bad out of date ad-hoc optimization techniques.

This isn’t meant to be a left handed compliment: algorithms are a first love of mine and some of the matching algorithms bioinformaticians uses (like online suffix trees) are quite brilliant.

variety of techniques from statistics, modeling, machine learning, and data mining that analyze current and historical facts to make predictions about future, or otherwise unknown, events.”

don’t tend to use the term predictive analytics because I come from a probability, simulation, algorithms and machine learning background and not from an analytics background.

The Wikipedia defines data science as a field that “incorporates varying elements and builds on techniques and theories from many fields, including math, statistics, data engineering, pattern recognition and learning, advanced computing, visualization, uncertainty modeling, data warehousing, and high performance computing with the goal of extracting meaning from data and creating data products.”

Data science is a term I use to represent the ownership and management of the entire modeling process: discovering the true business need, collecting data, managing data, building models and deploying models into production.

Machine learning

Machine learning is a field of computer science that often uses statistical techniques to give computers the ability to 'learn' (i.e., progressively improve performance on a specific task) with data, without being explicitly programmed.[1] The name machine learning was coined in 1959 by Arthur Samuel.[2] Evolved from the study of pattern recognition and computational learning theory in artificial intelligence,[3] machine learning explores the study and construction of algorithms that can learn from and make predictions on data[4] – such algorithms overcome following strictly static program instructions by making data-driven predictions or decisions,[5]:2 through building a model from sample inputs.

Mitchell provided a widely quoted, more formal definition of the algorithms studied in the machine learning field: 'A computer program is said to learn from experience E with respect to some class of tasks T and performance measure P if its performance at tasks in T, as measured by P, improves with experience E.'[13] This definition of the tasks in which machine learning is concerned offers a fundamentally operational definition rather than defining the field in cognitive terms.

Machine learning tasks are typically classified into two broad categories, depending on whether there is a learning 'signal' or 'feedback' available to a learning system: Another categorization of machine learning tasks arises when one considers the desired output of a machine-learned system:[5]:3 Among other categories of machine learning problems, learning to learn learns its own inductive bias based on previous experience.

Developmental learning, elaborated for robot learning, generates its own sequences (also called curriculum) of learning situations to cumulatively acquire repertoires of novel skills through autonomous self-exploration and social interaction with human teachers and using guidance mechanisms such as active learning, maturation, motor synergies, and imitation.

Probabilistic systems were plagued by theoretical and practical problems of data acquisition and representation.[17]:488 By 1980, expert systems had come to dominate AI, and statistics was out of favor.[18] Work on symbolic/knowledge-based learning did continue within AI, leading to inductive logic programming, but the more statistical line of research was now outside the field of AI proper, in pattern recognition and information retrieval.[17]:708–710;

Machine learning and data mining often employ the same methods and overlap significantly, but while machine learning focuses on prediction, based on known properties learned from the training data, data mining focuses on the discovery of (previously) unknown properties in the data (this is the analysis step of knowledge discovery in databases).

Much of the confusion between these two research communities (which do often have separate conferences and separate journals, ECML PKDD being a major exception) comes from the basic assumptions they work with: in machine learning, performance is usually evaluated with respect to the ability to reproduce known knowledge, while in knowledge discovery and data mining (KDD) the key task is the discovery of previously unknown knowledge.

Jordan, the ideas of machine learning, from methodological principles to theoretical tools, have had a long pre-history in statistics.[20] He also suggested the term data science as a placeholder to call the overall field.[20] Leo Breiman distinguished two statistical modelling paradigms: data model and algorithmic model,[21] wherein 'algorithmic model' means more or less the machine learning algorithms like Random forest.

Multilinear subspace learning algorithms aim to learn low-dimensional representations directly from tensor representations for multidimensional data, without reshaping them into (high-dimensional) vectors.[27] Deep learning algorithms discover multiple levels of representation, or a hierarchy of features, with higher-level, more abstract features defined in terms of (or generating) lower-level features.

In machine learning, genetic algorithms found some uses in the 1980s and 1990s.[31][32] Conversely, machine learning techniques have been used to improve the performance of genetic and evolutionary algorithms.[33] Rule-based machine learning is a general term for any machine learning method that identifies, learns, or evolves 'rules' to store, manipulate or apply, knowledge.

They seek to identify a set of context-dependent rules that collectively store and apply knowledge in a piecewise manner in order to make predictions.[35] Applications for machine learning include: In 2006, the online movie company Netflix held the first 'Netflix Prize' competition to find a program to better predict user preferences and improve the accuracy on its existing Cinematch movie recommendation algorithm by at least 10%.

A joint team made up of researchers from AT&T Labs-Research in collaboration with the teams Big Chaos and Pragmatic Theory built an ensemble model to win the Grand Prize in 2009 for $1 million.[41] Shortly after the prize was awarded, Netflix realized that viewers' ratings were not the best indicators of their viewing patterns ('everything is a recommendation') and they changed their recommendation engine accordingly.[42] In 2010 The Wall Street Journal wrote about the firm Rebellion Research and their use of Machine Learning to predict the financial crisis.

[43] In 2012, co-founder of Sun Microsystems Vinod Khosla predicted that 80% of medical doctors jobs would be lost in the next two decades to automated machine learning medical diagnostic software.[44] In 2014, it has been reported that a machine learning algorithm has been applied in Art History to study fine art paintings, and that it may have revealed previously unrecognized influences between artists.[45] Although machine learning has been very transformative in some fields, effective machine learning is difficult because finding patterns is hard and often not enough training data are available;

as a result, machine-learning programs often fail to deliver.[46][47] Classification machine learning models can be validated by accuracy estimation techniques like the Holdout method, which splits the data in a training and test set (conventionally 2/3 training set and 1/3 test set designation) and evaluates the performance of the training model on the test set.

Systems which are trained on datasets collected with biases may exhibit these biases upon use (algorithmic bias), thus digitizing cultural prejudices.[50] For example, using job hiring data from a firm with racist hiring policies may lead to a machine learning system duplicating the bias by scoring job applicants against similarity to previous successful applicants.[51][52] Responsible collection of data and documentation of algorithmic rules used by a system thus is a critical part of machine learning.

There is huge potential for machine learning in health care to provide professionals a great tool to diagnose, medicate, and even plan recovery paths for patients, but this will not happen until the personal biases mentioned previously, and these 'greed' biases are addressed.[54] Software suites containing a variety of machine learning algorithms include the following :

Difference between Machine Learning & Statistical Modeling

One of the most common question, which gets asked at various data science forums is: What is the difference between Machine Learning and Statistical modeling?

When I came across this question at first, I found almost no clear answer which can layout how machine learning is different from statistical modeling.

Given the similarity in terms of the objective both try to solve for, the only difference lies in the volume of data involved and human involvement for building a model.

Here is an interesting Venn diagram on the coverage of machine learning and statistical modeling in the universe of data science (Reference: SAS institute)

The common objective behind using either of the tools is Learning from Data. Both these approaches aim to learn about the underlying phenomena by using data generated in the process.

Let us now see an interesting example published by McKinsey differentiating the two algorithms : Case : Understand the risk level of customers churn over a period of time for a Telecom company Data Available : Two Drivers –

Even with a laptop of 16 GB RAM I daily work on datasets of millions of rows with thousands of parameter and build an entire model in not more than 30 minutes.

Given the flavor of difference in output of these two approaches, let us understand the difference in the two paradigms, even though both do almost similar job : All the differences mentioned above do separate the two to some extent, but there is no hard boundary between Machine Learning and statistical modeling.

subfield of computer science and artificial intelligence which deals with building systems that can learn from data, instead of explicitly programmed instructions.

It came into existence in the 1990s as steady advances in digitization and cheap computing power enabled data scientists to stop building finished models and instead train computers to do so.

The unmanageable volume and complexity of the big data that the world is now swimming in have increased the potential of machine learning—and the need for it.

In a statistical model, we basically try to estimate the function f in Machine Learning takes away the deterministic function “f”

It simply becomes It will try to find pockets of X in n dimensions (where n is the number of attributes), where occurrence of Y is significantly different.


In statistics, overfitting is 'the production of an analysis that corresponds too closely or exactly to a particular set of data, and may therefore fail to fit additional data or predict future observations reliably'.[1] An overfitted model is a statistical model that contains more parameters than can be justified by the data.[2] The essence of overfitting is to have unknowingly extracted some of the residual variation (i.e.

The potential for overfitting depends not only on the number of parameters and data but also the conformability of the model structure with the data shape, and the magnitude of model error compared to the expected level of noise or error in the data.[citation needed] Even when the fitted model does not have an excessive number of parameters, it is to be expected that the fitted relationship will appear to perform less well on a new data set than on the data set used for fitting (a phenomenon sometimes known as shrinkage).[2] In particular, the value of the coefficient of determination will shrink relative to the original data.

The basis of some techniques is either (1) to explicitly penalize overly complex models or (2) to test the model's ability to generalize by evaluating its performance on a set of data not used for training, which is assumed to approximate the typical unseen data that a model will encounter.

Anderson, in their much-cited text on model selection, argue that to avoid overfitting, we should adhere to the 'Principle of Parsimony'.[3] The authors also state the following.[3]:32-33 Overfitted models … are often free of bias in the parameter estimators, but have estimated (and actual) sampling variances that are needlessly large (the precision of the estimators is poor, relative to what could have been accomplished with a more parsimonious model).

In regression analysis, overfitting occurs frequently.[5] In the extreme case, if there are p variables in a linear regression with p data points, the fitted hyperplane will go exactly through every point.[6] A study in 2015 suggests that two observations per independent variable are sufficient for linear regression[7].

In the process of regression model selection, the mean squared error of the random regression function can be split into random noise, approximation bias, and variance in the estimate of the regression function, and the bias–variance tradeoff is often used to overcome overfit models.

If the new, more complicated function is selected instead of the simple function, and if there was not a large enough gain in training-data fit to offset the complexity increase, then the new complex function 'overfits' the data, and the complex overfitted function will likely perform worse than the simpler function on validation data outside the training dataset, even though the complex function performed as well, or perhaps even better, on the training dataset.[11] When comparing different types of models, complexity cannot be measured solely by counting how many parameters exist in each model;

For example, it is nontrivial to directly compare the complexity of a neural net (which can track curvilinear relationships) with m parameters to a regression model with n parameters.[11] Overfitting is especially likely in cases where learning was performed too long or where training examples are rare, causing the learner to adjust to very specific random features of the training data, that have no causal relation to the target function.

Glossary of common Machine Learning, Statistics and Data Science terms

In the coming days, we will add more terms related to data science, business intelligence and big data.

In the meanwhile, if you want to contribute to the glossary or want to request adding more terms, please feel free to let us know through comments below!

When we get an output as probabilities and have to classify them into classes, we decide some threshold value and if the probability is above that threshold value we classify it as 1, and 0 otherwise.

Vision is a field of computer science that deals with enabling computers to visualize, process and identify images/videos in the same way that a human vision does.

In the recent times, the major driving forces behind Computer Vision has been the emergence of deep learning, rise in computational power and a huge amount of image data.

The image data can take many forms, such as video sequences, views from multiple cameras, or multi-dimensional data from a medical scanner.

The number of concordant and discordant pairs are used in calculations for Kendall’s tau, which measures the association between two ordinal variables. Let’s

ranks given by the reviewer 1 are ordered in ascending order, this way we can compare the rankings given by both the reviewers.

confidence interval is used to estimate what percent of a population fits a category based on the results from a sample population.

For example, if 70 adults own a cell phone in a random sample of 100 adults, we can be fairly confident that the true percentage amongst the population is somewhere between 61% and 79%.

The 2nd quadrant is called type II error or False Negatives, whereas 3rd quadrant is called type I error or False positives Continuous

An iterative algorithm is said to converge when as the iterations proceed the output gets closer and closer to a specific value. Convex

real value function is called convex if the line segment between any two points on the graph of the function lies above or on the graph. Convex

Two parallel vectors have a cosine similarity of 1 and two vectors at 90° have a cosine similarity of 0.

Suppose we have two vectors A and B, cosine similarity of these vectors can be calculated by dividing the dot product of A and B with the product of the magnitude of the two vectors.

It’s similar to variance, but where variance tells you how a single variable varies, co variance tells you how two variables vary together.

positive covariance means the variables are positively related, while a negative covariance means the variables are inversely related. Cross

information theory, the cross entropy between two probability distributions and over the same underlying set of events measures the average number of bits needed to identify an event drawn from the set, if a coding scheme is used that is optimized for an “unnatural”

stopping is a technique for avoiding overfitting when training a machine learning model with iterative method.

We set the early stopping in such a way that when the performance has stopped improving on the held-out validation set, the model training stops.

Early stopping enables you to specify a validation dataset and the number of iterations after which the algorithm should stop if the score on your validation dataset didn’t increase. EDA EDA

or exploratory data analysis is a phase used for data science pipeline in which the focus is to understand insights of the data through visualization or by statistical analysis. The

enforces data quality and consistency standards Delivers data in a presentation-ready format This data can be used by application developers to build applications and end users for making decisions. Evaluation

Without looking up the indices in an associative array, it applies a hash function to the features and uses their hash values as indices directly.

Selection is a process of choosing those features which are required to explain the predictive power of a statistical model and dropping out irrelevant features. This

learning refers to the training of machine learning algorithms using a very small set of training data instead of a very large set.

This is most suitable in the field of computer vision, where it is desirable to have an object categorization model work well without thousands of training examples. Flume Flume

It calculates the probability of an event in the long run of the experiment (i.e the experiment is repeated under the same conditions to obtain the outcome). Here,

For example, I perform an experiment with a stopping intention in mind that I will stop the experiment when it is repeated 1000 times or I see minimum 300 heads in a coin toss.

(or interquartile range) is a measure of variability based on dividing the rank-ordered data set into four equal parts.

For example, each iteration of training a neural network takes certain number of training data and updates the weights by using gradient descent or some other weight update rule.

data are usually more expensive to obtain than the raw unlabeled data because preparation of the labelled data involves manual labelling every piece of unlabeled data. Labeled

These charts are used to communicate information visually, such as to show an increase or decrease in the trend in data over intervals of time.

In the plot below, for each time instance, the speed trend is shown and the points are connected to display the trend over time. This

Line charts can also be used to compare changes over the same period of time for multiple cases, like plotting the speed of a cycle, car, train over time in the same plot. Linear

Let us say, you ask a child in fifth grade to arrange people in his class by increasing order of weight, without asking them their weight!

He / she would likely look (visually analyze) at the height and build of people and arrange them using a combination of these visible parameters.

coefficients a and b are derived based on minimizing the sum of squared difference of distance between data points and regression line. Look

simple words, it predicts the probability of occurrence of an event by fitting data to a logistic function. Hence, it is also known as logistic regression.

artificial neuron, as in a multi-layer neural network, that is, they compute an activation (using an activation function) of a weighted sum.

is a classification technique based on Bayes’ theorem with an assumption of independence between predictors. In simple terms, a Naive Bayes classifier assumes that the presence of a particular feature in a class is unrelated to the presence of any other feature.

Even if these features depend on each other or upon the existence of the other features, a naive Bayes classifier would consider all of these properties to independently contribute to the probability that this fruit is an apple. NaN NaN

A NoSQL database provides a mechanism for storage and retrieval of data that is modeled in means other than the tabular relations used in relational databases.

It can accommodate a wide variety of data models, including key-value, document, columnar and graph formats.

example, if you have a variable ranging from 0 to 1 and other from 0 to 1000, you can normalize the variable, such that both are in the range 0 to 1. Numpy NumPy

PWordDescriptionPandasPandas is an open source, high-performance, easy-to-use data structure and data analysis library for the Python programming language.

for reading and writing data between in-memory data structures and different formats: CSV and text files, Microsoft Excel, SQL databases, and the fast HDF5 format. Flexible

computer vision, supervised pattern recognition techniques are used for optical character recognition (OCR), face detection, face recognition, object detection, and object classification. Pie

You can spend years to build a decent image recognition algorithm from scratch or you can take inception model (a pre-trained model) from Google which was built on ImageNet data to identify images in those pictures. For

is an open source programming language, widely used for various applications, such as general purpose programming, data science and machine learning.

RWordDescription RR is an open-source programming language and a software environment for statistical computing, machine learning, and data visualization.

Recommendation engines basically are data filtering tools that make use of algorithms and data to recommend the most relevant items to a particular user.

If we can recommend items to a customer based on their needs and interests, it will create a positive effect on the user experience and they will visit more frequently.

is an example of machine learning where the machine is trained to take specific decisions based on the business requirement with the sole motto to maximize efficiency (performance).

The idea involved in reinforcement learning is: The machine/ software agent trains itself on a continual basis based on the environment it is exposed to, and applies it’s enriched knowledge to solve business problems.

An RL agent learns from its past experience, rather from its continual trial and error learning process as against supervised learning where an external supervisor  provides examples. A

Self driving cars use Reinforcement learning to make decisions continuously like which route to take, what speed to drive on, are some of the questions which are decided after interacting with the environment.

α (alpha) is the parameter which balances the amount of emphasis given to minimizing RSS vs minimizing sum of squares of coefficients.

mathematics, a function defined on an inner product space is said to have rotational invariance if its value does not change when arbitrary rotations are applied to its argument.

Find the Mean of set of numbers, which is (600 + 470 + 170 + 430 + 300) / 5 = 394 2)

Statistical modeling and machine learning

This lecture for students in bioinformatics, physics, master students of Data Engineering and Analytics, and master students of Biomedical Computing will lay the theoretical and practical foundations for statistical data analysis, statistical modeling and machine learning from a Bayesian probabilistic perspective.

It will train you in "thinking probabilistically", which is extremely useful in many areas of quantitative sciences where noisy data need to be analyzed (and hence modeled!).

The lecture will be held in inverted classroom style: Each week, we will give a ~30 min overview of the next reading assignment of a section of the book, pointing out the essential messages, thus facilitating the reading at home.

Then, practical exercises using the newly acquired material will be solved in teams, using the R statistics framework (100min).

The inverted classroom style is in our experience better suited than the conventional lecturing model for quantitative topics that require the students to think through or retrace mathematical derivations at their own speed.

Math Basic linear algebra (matrix and vector algebra, inverse and transposed matrices, determinants, eigenvalue decomposition) and one-dimensional calculus (e.g.

Probability theory, Bayes theorem, conditional independence, distributions (multinomial, Poisson, Gaussian, gamma, beta,...), central limit theorem, entropy, mutual information 3.

Bayesian statistics: max posterior estimation, model selection, uninformative and robust priors, hierarchical and empirical Bayes, Bayesian decision theory 6.

StatQuest: What is a statistical model?

"Model" is a vague term that means different things in different contexts. Here I clear it all up in the context of statistics! If you'd like to support StatQuest, please ...

Linear Regression - Machine Learning Fun and Easy

Linear Regression - Machine Learning Fun and Easy

Statistics 03: Types of statistical models

In this lecture, I show which types of statistical models should be used when; the most important decision concerns the explanatory variables: When these are ...

Principal Components Analysis - Georgia Tech - Machine Learning

Watch on Udacity: Check out the full Advanced Operating Systems course for free ..

Random Forest - Fun and Easy Machine Learning

Random Forest - Fun and Easy Machine Learning

Predicting the Winning Team with Machine Learning

Can we predict the outcome of a football game given a dataset of past games? That's the question that we'll answer in this episode by using the scikit-learn ...

Machine Learning and Predictive Analytics - Continuous and Categorical Features (Cardinality)

Machine Learning and Predictive Analytics. #MachineLearning Learn More: (Fundamentals Of Machine Learning for Predictive Data ..

Choosing which statistical test to use - statistics help

Seven different statistical tests and a process by which you can decide which to use. If this video helps you, please donate by clicking on: ...

Bag of Words - Intro to Machine Learning

This video is part of an online course, Intro to Machine Learning. Check out the course here: This course was designed ..

How SVM (Support Vector Machine) algorithm works

In this video I explain how SVM (Support Vector Machine) algorithm works to classify a linearly separable binary data set. The original presentation is available ...