You're reading from Data Forecasting and Segmentation Using Microsoft Excel Perform data grouping, linear predictions, and time series machine learning statistics without using code

Product type Paperback

Published in May 2022

Publisher Packt

ISBN-13 9781803247731

Length 324 pages

Edition 1st Edition

Languages

C++

Tools

Excel

Concepts

Data Analysis

Author (1):

Fernando Roque

View More author details

Table of Contents (19) Chapters

Preface

1. Part 1 – An Introduction to Machine Learning Functions

2. Chapter 1: Understanding Data Segmentation FREE CHAPTER

3. Chapter 2: Applying Linear Regression

4. Chapter 3: What is Time Series?

5. Part 2 – Grouping Data to Find Segments and Outliers

6. Chapter 4: Introduction to Data Grouping

7. Chapter 5: Finding the Optimal Number of Single Variable Groups

8. Chapter 6: Finding the Optimal Number of Multi-Variable Groups

9. Chapter 7: Analyzing Outliers for Data Anomalies

10. Part 3 – Simple and Multiple Linear Regression Analysis

11. Chapter 8: Finding the Relationship between Variables

12. Chapter 9: Building, Training, and Validating a Linear Model

13. Chapter 10: Building, Training, and Validating a Multiple Regression Model

14. Part 4 – Predicting Values with Time Series

15. Chapter 11: Testing Data for Time Series Compliance

16. Chapter 12: Working with Time Series Using the Centered Moving Average and a Trending Component

17. Chapter 13: Training, Validating, and Running the Model

18. Other Books You May Enjoy

K-means data grouping

Before running the K-means function, we need to calculate the parameter of the optimal number of groups. In the last section, we saw that the first approach is to visualize the 2D and 3D charts. However, the best way to get the number of groups is by calculating the elbow function and choosing the number of groups when the curve of the elbow starts to flatten.

Running the elbow algorithm

Now that we have an idea of the number of groups for the credit card fraud dataset, we are ready to use the elbow K-means algorithm to get a statistical value of the optimal number of groups.

Kaggle credit card fraud dataset

Before we run the K-means function, we have to calculate the optimal number of groups for the V1, Time (seconds), and Amount fields of the credit card transactions.

We run the elbow algorithm, changing the range of data in line 3 of the kmeanselbow.r add-in, as you can see in Figure 7.4. Use the credicard01.xlsx file included in this chapter...