Packt+ | Advance your knowledge in tech

You're reading from Scala Machine Learning Projects Build real-world machine learning and deep learning projects with Scala

Product type Paperback

Published in Jan 2018

Publisher Packt

ISBN-13 9781788479042

Length 470 pages

Edition 1st Edition

Languages

Scala

Tools

Apache Spark

Concepts

Deep Learning

Author (1):

Md. Rezaul Karim

View More author details

Table of Contents (13) Chapters

Preface

1. Analyzing Insurance Severity Claims

2. Analyzing and Predicting Telecommunication Churn FREE CHAPTER

3. High Frequency Bitcoin Price Prediction from Historical and Live Data

4. Population-Scale Clustering and Ethnicity Prediction

5. Topic Modeling - A Better Insight into Large-Scale Texts

6. Developing Model-based Movie Recommendation Engines

7. Options Trading Using Q-learning and Scala Play Framework

8. Clients Subscription Assessment for Bank Telemarketing using Deep Neural Networks

9. Fraud Analytics Using Autoencoders and Anomaly Detection

10. Human Activity Recognition using Recurrent Neural Networks

11. Image Classification using Convolutional Neural Networks

12. Other Books You May Enjoy

Leave a review - let other readers know what you think

Developing a fraud analytics model

Before we fully start, we need to do two things: know the dataset, and then prepare our programming environment.

Description of the dataset and using linear models

For this project, we will be using the credit card fraud detection dataset from Kaggle. The dataset can be downloaded from https://www.kaggle.com/dalpozz/creditcardfraud. Since I am using the dataset, it would be a good idea to be transparent by citing the following publication:

Andrea Dal Pozzolo, Olivier Caelen, Reid A. Johnson, and Gianluca Bontempi, Calibrating Probability with Undersampling for Unbalanced Classification. In Symposium on Computational Intelligence and Data Mining (CIDM), IEEE, 2015.

The datasets contain transactions made by credit cards by European cardholders in September 2013 over the span of only two days. There is a total of 285,299 transactions, with only 492 frauds out of 284,807 transactions, meaning the dataset is highly imbalanced and the positive class (fraud) accounts...