Book Image

Data Science Using Python and R

By : Chantal D. Larose, Daniel T. Larose
Book Image

Data Science Using Python and R

By: Chantal D. Larose, Daniel T. Larose

Overview of this book

Data science is hot. Bloomberg named a data scientist as the ‘hottest job in America’. Python and R are the top two open-source data science tools using which you can produce hands-on solutions to real-world business problems, using state-of-the-art techniques. Each chapter in the book presents step-by-step instructions and walkthroughs for solving data science problems using Python and R. You’ll learn how to prepare data, perform exploratory data analysis, and prepare to model the data. As you progress, you’ll explore what are decision trees and how to use them. You’ll also learn about model evaluation, misclassification costs, naïve Bayes classification, and neural networks. The later chapters provide comprehensive information about clustering, regression modeling, dimension reduction, and association rules mining. The book also throws light on exciting new topics, such as random forests and general linear models. The book emphasizes data-driven error costs to enhance profitability, which avoids the common pitfalls that may cost a company millions of dollars. By the end of this book, you’ll have enough knowledge and confidence to start providing solutions to data science problems using R and Python.
Table of Contents (20 chapters)
Free Chapter
1
ABOUT THE AUTHORS
17
INDEX
18
END USER LICENSE AGREEMENT

14.7 THE CONFIDENCE QUOTIENT CRITERION

To demonstrate the confidence quotient criterion, we generate rules using a lower bound of 40 (along with minimum antecedent support of 1%, minimum rule confidence of 5%, and maximum antecedents of 1). After removing any rules with Churn in the antecedent, we obtain the three association rules shown in Figure 14.7.

No alt text required.

Figure 14.7 Association rules in R found using confidence quotient lower bound = 40.

The confidence quotient evaluation measure gives the absolute ratio between the prior probability of the consequent (Churn = True) and the confidence of the rule. So, rules would be included in this case only if:

equation{1Rule confidencePrior proportion ofconsequent0.40or1Prior proportion of consequentRule confidence0.40--

whichever is not negative.

Let us confirm the calculations for Rule [3] from Figure 14.5. This rule has confidence 0.43051. From Figure 14.1, the prior proportion of the consequent...