Data Science

Concepts and Practice

Vijay Kotu, Bala Deshpande

• Verlag: Morgan Kaufmann
• Jahr: 2018
• ISBN:
  9780128147627
• Auflage: 2
• Thema: Informatik
• Thema: Künstliche Intelligenz (KI) & Semantik

Learn the basics of Data Science through an easy to understand conceptual framework and immediately practice using RapidMiner platform. Whether you are brand new to data science or working on your tenth project, this book will show you how to analyze data, uncover hidden patterns and relationships to aid important decisions and predictions.

Data Science has become an essential tool to extract value from data for any organization that collects, stores and processes data as part of its operations. This book is ideal for business users, data analysts, business analysts, engineers, and analytics professionals and for anyone who works with data.

You'll be able to:

• Gain the necessary knowledge of different data science techniques to extract value from data.

• Master the concepts and inner workings of 30 commonly used powerful data science algorithms.

• Implement step-by-step data science process using using RapidMiner, an open source GUI based data science platform

Data Science techniques covered: Exploratory data analysis, Visualization, Decision trees, Rule induction, k-nearest neighbors, Naïve Bayesian classifiers, Artificial neural networks, Deep learning, Support vector machines, Ensemble models, Random forests, Regression, Recommendation engines, Association analysis, K-Means and Density based clustering, Self organizing maps, Text mining, Time series forecasting, Anomaly detection, Feature selection and more...

• Contains fully updated content on data science, including tactics on how to mine business data for information
• Presents simple explanations for over twenty powerful data science techniques
• Enables the practical use of data science algorithms without the need for programming
• Demonstrates processes with practical use cases
• Introduces each algorithm or technique and explains the workings of a data science algorithm in plain language
• Describes the commonly used setup options for the open source tool RapidMiner

Chapter 1

Introduction

Abstract

Data science has been growing in popularity over the past few years. In the introduction, the terms “data science” and its taxonomy are defined. This chapter covers the relationship between artificial intelligence, machine learning, and data science, provides the motivation for data science, an introduction to key algorithms, and presents a roadmap for rest of the book.

Keywords

Data science; descriptive analytics; taxonomy; classification; regression; association; clustering; anomaly; RapidMiner; roadmap

Data science is a collection of techniques used to extract value from data. It has become an essential tool for any organization that collects, stores, and processes data as part of its operations. Data science techniques rely on finding useful patterns, connections, and relationships within data. Being a buzzword, there is a wide variety of definitions and criteria for what constitutes data science. Data science is also commonly referred to as knowledge discovery, machine learning, predictive analytics, and data mining. However, each term has a slightly different connotation depending on the context. In this chapter, we attempt to provide a general overview of data science and point out its important features, purpose, taxonomy, and methods.

In spite of the present growth and popularity, the underlying methods of data science are decades if not centuries old. Engineers and scientists have been using predictive models since the beginning of nineteenth century. Humans have always been forward-looking creatures and predictive sciences are manifestations of this curiosity. So, who uses data science today? Almost every organization and business. Sure, we didn’t call the methods that are now under data science as “Data Science.” The use of the term science in data science indicates that the methods are evidence based, and are built on empirical knowledge, more specifically historical observations.

As the ability to collect, store, and process data has increased, in line with Moore’s Law - which implies that computing hardware capabilities double every two years, data science has found increasing applications in many diverse fields. Just decades ago, building a production quality regression model took about several dozen hours (Parr Rud, 2001). Technology has come a long way. Today, sophisticated machine learning models can be run, involving hundreds of predictors with millions of records in a matter of a few seconds on a laptop computer.

The process involved in data science, however, has not changed since those early days and is not likely to change much in the foreseeable future. To get meaningful results from any data, a major effort preparing, cleaning, scrubbing, or standardizing the data is still required, before the learning algorithms can begin to crunch them. But what may change is the automation available to do this. While today this process is iterative and requires analysts’ awareness of the best practices, soon smart automation may be deployed. This will allow the focus to be put on the most important aspect of data science: interpreting the results of the analysis in order to make decisions. This will also increase the reach of data science to a wider audience.

When it comes to the data science techniques, are there a core set of procedures and principles one must master? It turns out that a vast majority of data science practitioners today use a handful of very powerful techniques to accomplish their objectives: decision trees, regression models, deep learning, and clustering (Rexer, 2013). A majority of the data science activity can be accomplished using relatively few techniques. However, as with all 80/20 rules, the long tail, which is made up of a large number of specialized techniques, is where the value lies, and depending on what is needed, the best approach may be a relatively obscure technique or a combination of several not so commonly used procedures. Thus, it will pay off to learn data science and its methods in a systematic way, and that is what is covered in these chapters. But, first, how are the often-used terms artificial intelligence (AI), machine learning, and data science explained?

1.1 AI, Machine learning, and Data Science

Artificial intelligence, Machine learning, and data science are all related to each other. Unsurprisingly, they are often used interchangeably and conflated with each other in popular media and business communication. However, all of these three fields are distinct depending on the context. Fig. 1.1 shows the relationship between artificial intelligence, machine learning, and data science.

Figure 1.1 Artificial intelligence, machine learning, and data science.

Artificial intelligence is about giving machines the capability of mimicking human behavior, particularly cognitive functions. Examples would be: facial recognition, automated driving, sorting mail based on postal code. In some cases, machines have far exceeded human capabilities (sorting thousands of postal mails in seconds) and in other cases we have barely scratched the surface (search “artificial stupidity”). There are quite a range of techniques that fall under artificial intelligence: linguistics, natural language processing, decision science, bias, vision, robotics, planning, etc. Learning is an important part of human capability. In fact, many other living organisms can learn.

Machine learning can either be considered a sub-field or one of the tools of artificial intelligence, is providing machines with the capability of learning from experience. Experience for machines comes in the form of data. Data that is used to teach machines is called training data. Machine learning turns the traditional programing model upside down (Fig. 1.2). A program, a set of instructions to a computer, transforms input signals into output signals using predetermined rules and relationships. Machine learning algorithms, also called “learners”, take both the known input and output (training data) to figure out a model for the program which converts input to output. For example, many organizations like social media platforms, review sites, or forums are required to moderate posts and remove abusive content. How can machines be taught to automate the removal of abusive content? The machines need to be shown examples of both abusive and non-abusive posts with a clear indication of which one is abusive. The learners will generalize a pattern based on certain words or sequences of words in order to conclude whether the overall post is abusive or not. The model can take the form of a set of “if–-then” rules. Once the data science rules or model is developed, machines can start categorizing the disposition of any new posts.

Figure 1.2 Traditional program and machine learning.

Data science is the business application of machine learning, artificial intelligence, and other quantitative fields like statistics, visualization, and mathematics. It is an interdisciplinary field that extracts value from data. In the context of how data science is used today, it relies heavily on machine learning and is sometimes called data mining. Examples of data science user cases are: recommendation engines that can recommend movies for a particular user, a fraud alert model that detects fraudulent credit card transactions, find customers who will most likely churn next month, or predict revenue for the next quarter.

1.2 What is Data Science?

Data science starts with data, which can range from a simple array of a few numeric observations to a complex matrix of millions of observations with thousands of variables. Data science utilizes certain specialized computational methods in order to discover meaningful and useful structures within a dataset. The discipline of data science coexists and is closely associated with a number of related areas such as database systems, data engineering, visualization, data analysis, experimentation, and business intelligence (BI). We can further define data science by investigating some of its key features and motivations.

1.2.1 Extracting Meaningful Patterns

Knowledge discovery in databases is the nontrivial process of identifying valid, novel, potentially useful, and ultimately understandable patterns or relationships within a dataset in order to make important decisions (Fayyad, Piatetsky-shapiro, & Smyth, 1996). Data science involves inference and iteration of many different hypotheses. One of the key aspects of data science is the process of generalization of patterns from a dataset. The generalization should be valid, not just for the dataset used to observe the pattern, but also for new unseen data. Data science is also a process with defined steps, each with a set of tasks. The term novel indicates that data science is usually involved in finding previously unknown patterns in data. The ultimate objective of data science is to find potentially useful conclusions that can be acted upon by the users of the analysis.

1.2.2 Building Representative Models

In statistics, a model is the representation of a relationship between variables in a dataset. It describes how one or more variables in the data are related to other variables. Modeling is a process in which a representative abstraction is built from the observed dataset. For example, based on credit score, income level, and requested loan amount, a model can be developed to determine the interest rate of a loan. For this task, previously known observational data including credit score, income level, loan amount, and interest rate are needed. Fig. 1.3 shows the process of generating a model. Once the representative model is created, it can be used to predict the value of the interest rate, based on all the input variables.

Figure 1.3 Data science models.

Data science is the process of building a representative model that fits the observational data. This model serves two purposes: on the one hand, it predicts the output (interest rate) based on the new and unseen set of input variables (credit score, income level, and loan amount), and on the other hand, the model can be used to understand the relationship between the output variable and all the input variables. For example, does income level really matter in determining the interest rate of a loan? Does income level matter more than credit score? What happens when income levels double or if credit score drops by 10 points? A Model can be used for both predictive and explanatory applications.

1.2.3 Combination of Statistics, Machine Learning, and Computing

In the pursuit of extracting useful and relevant information from large datasets, data science borrows computational techniques from the disciplines of statistics, machine learning, experimentation, and database theories. The algorithms used in data science originate from these disciplines but have since evolved to adopt more diverse techniques such as parallel computing, evolutionary computing, linguistics, and behavioral studies. One of the key ingredients of successful data science is substantial prior knowledge about the data and the business processes that generate the data, known as subject matter expertise. Like many quantitative frameworks, data science is an iterative process in which the practitioner gains more information about the patterns and relationships from data in each cycle. Data science also typically operates on large datasets that need to be stored, processed, and computed. This is where database techniques along with parallel and distributed computing techniques play an important role in data science.

1.2.4 Learning Algorithms

We can also define data science as a process of discovering previously unknown patterns in data using automatic iterative methods. The application of sophisticated learning algorithms for extracting useful patterns from data differentiates data science from traditional data analysis techniques. Many of these algorithms were developed in the past few decades and are a part of machine learning and artificial intelligence. Some algorithms are based on the foundations of Bayesian probabilistic theories and regression analysis, originating...