Drowning in Data: A Modern Data Analyst's Survival Guide

This book illustrates how critical thinking and classical statistical techniques can be used to solve big data problems.  Through fun, real-world case studies, the book also shows how to check data integrity, manage large datasets, validate models, and characterize uncertainty.  Common methods of modern data analytics, including dimension reduction, classification and clustering, and trend detection, are presented within chapters that focus on specific classes of techniques. Common methods are discussed in mathematical detail, but rather than focusing on these methods, the book emphasizes their application to real-world data analysis. For example, a chapter is devoted to supervised classification techniques. In this chapter, the k-nearest neighbor, linear discrimination, and classification tree algorithms are presented. Each method is applied to a case study, and the data collection and cleaning process is illustrated.  Another chapter focuses on data integration methods, which are used to integrate the disparate data types, and cross-validation is used to measure model accuracy. This approach brings together many components of modern data analytics, allowing for the creative application of statistical techniques rather than the techniques themselves.