Book Image

Hands-On Machine Learning for Cybersecurity

By : Soma Halder, Sinan Ozdemir
Book Image

Hands-On Machine Learning for Cybersecurity

By: Soma Halder, Sinan Ozdemir

Overview of this book

Cyber threats today are one of the costliest losses that an organization can face. In this book, we use the most efficient tool to solve the big problems that exist in the cybersecurity domain. The book begins by giving you the basics of ML in cybersecurity using Python and its libraries. You will explore various ML domains (such as time series analysis and ensemble modeling) to get your foundations right. You will implement various examples such as building system to identify malicious URLs, and building a program to detect fraudulent emails and spam. Later, you will learn how to make effective use of K-means algorithm to develop a solution to detect and alert you to any malicious activity in the network. Also learn how to implement biometrics and fingerprint to validate whether the user is a legitimate user or not. Finally, you will see how we change the game with TensorFlow and learn how deep learning is effective for creating models and training systems
Table of Contents (13 chapters)
Free Chapter
Basics of Machine Learning in Cybersecurity
Using Data Science to Catch Email Fraud and Spam

Logistic regression classifier – under-sampled data

We are interested in the recall score, because that is the metric that will help us try to capture the most fraudulent transactions. If you think how accuracy, precision, and recall work for a confusion matrix, recall would be the most interesting because we comprehend a lot more.

  • Accuracy = (TP+TN)/total, where TP depicts true positive, TN depicts true negative
  • Precision = TP/(TP+FP), where TP depicts true positive, FP depicts false positive
  • Recall = TP/(TP+FN), where TP depicts true positive, TP depicts true positive, FN depicts false negative

The following diagram will help you understand the preceding definitions:

As we know, due to the imbalance of data, many observations could be predicted as False Negatives. However, in our case, that is not so; we do not predict a normal transaction. The transaction is in fact...