Packt+ | Advance your knowledge in tech

You're reading from TensorFlow Machine Learning Projects Build 13 real-world projects with advanced numerical computations using the Python ecosystem

Product type Paperback

Published in Nov 2018

Publisher Packt

ISBN-13 9781789132212

Length 322 pages

Edition 1st Edition

Languages

Python

Tools

TensorFlow

Concepts

Machine Learning

Authors (2):

Ankit Jain

Dr. Amita Kapoor

View More author details

Table of Contents (17) Chapters

Preface

1. Overview of TensorFlow and Machine Learning FREE CHAPTER

2. Using Machine Learning to Detect Exoplanets in Outer Space

3. Sentiment Analysis in Your Browser Using TensorFlow.js

4. Digit Classification Using TensorFlow Lite

5. Speech to Text and Topic Extraction Using NLP

6. Predicting Stock Prices using Gaussian Process Regression

7. Credit Card Fraud Detection using Autoencoders

8. Generating Uncertainty in Traffic Signs Classifier Using Bayesian Neural Networks

9. Generating Matching Shoe Bags from Shoe Images Using DiscoGANs

10. Classifying Clothing Images using Capsule Networks

11. Making Quality Product Recommendations Using TensorFlow

12. Object Detection at a Large Scale with TensorFlow

13. Generating Book Scripts Using LSTMs

14. Playing Pacman Using Deep Reinforcement Learning

15. What is Next?

16. Other Books You May Enjoy

Leave a review - let other readers know what you think

Building a fraud detection model

For this project, we are going to use the credit card dataset from Kaggle (https://www.kaggle.com/mlg-ulb/creditcardfraud), 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. It consists of credit card transaction data from two days, from European cardholders. The dataset is highly imbalanced and contains approximately 284,000 pieces of transaction data with 492 instances of fraud (0.172% of the total).

There are 31 numerical columns in the dataset. Two of them are time and amount. Time denotes the amount of time elapsed (in seconds) between each transaction and the first transaction in the dataset. Amount is the total amount regarding the transaction. For our model, we will eliminate the time column as it doesn't help with the accuracy of the model. The rest of the features (V1, V2...