Book Image

Statistics for Data Science

Book Image

Statistics for Data Science

Overview of this book

Data science is an ever-evolving field, which is growing in popularity at an exponential rate. Data science includes techniques and theories extracted from the fields of statistics; computer science, and, most importantly, machine learning, databases, data visualization, and so on. This book takes you through an entire journey of statistics, from knowing very little to becoming comfortable in using various statistical methods for data science tasks. It starts off with simple statistics and then move on to statistical methods that are used in data science algorithms. The R programs for statistical computation are clearly explained along with logic. You will come across various mathematical concepts, such as variance, standard deviation, probability, matrix calculations, and more. You will learn only what is required to implement statistics in data science tasks such as data cleaning, mining, and analysis. You will learn the statistical techniques required to perform tasks such as linear regression, regularization, model assessment, boosting, SVMs, and working with neural networks. By the end of the book, you will be comfortable with performing various statistical computations for data science programmatically.
Table of Contents (19 chapters)
Title Page
Credits
About the Author
About the Reviewer
www.PacktPub.com
Customer Feedback
Preface

Assessment and statistical assessment


Merriam-Webster definesassessment as:

The action or an instance of making a judgment about something.

The following image shows flow for assessing statistical data:

We need to keep a few pointers in mind for statistical assessment. They are listed as follows.

Objectives

With that in mind, to be able to make a reasonable assessment--that is, make a judgment--on something (anything really), one must first have to set objective(s). Assessment objectives help the data scientist determine how to assess data, a database, or a statistical data model. Without clear objectives, you'll waste valuable time, and potentially, put confidence in a model that doesn't meet a business requirement or may even lead to incorrect assumptions (predictions).

Baselines

Next, (based on your set objectives) standards, minimum acceptable performance, or a baseline to establish an opinion on what is being assessed need to be established. In other words, how well does what you are assessing...