Book Image

Python Machine Learning Blueprints - Second Edition

By : Alexander Combs, Michael Roman
Book Image

Python Machine Learning Blueprints - Second Edition

By: Alexander Combs, Michael Roman

Overview of this book

Machine learning is transforming the way we understand and interact with the world around us. This book is the perfect guide for you to put your knowledge and skills into practice and use the Python ecosystem to cover key domains in machine learning. This second edition covers a range of libraries from the Python ecosystem, including TensorFlow and Keras, to help you implement real-world machine learning projects. The book begins by giving you an overview of machine learning with Python. With the help of complex datasets and optimized techniques, you’ll go on to understand how to apply advanced concepts and popular machine learning algorithms to real-world projects. Next, you’ll cover projects from domains such as predictive analytics to analyze the stock market and recommendation systems for GitHub repositories. In addition to this, you’ll also work on projects from the NLP domain to create a custom news feed using frameworks such as scikit-learn, TensorFlow, and Keras. Following this, you’ll learn how to build an advanced chatbot, and scale things up using PySpark. In the concluding chapters, you can look forward to exciting insights into deep learning and you'll even create an application using computer vision and neural networks. By the end of this book, you’ll be able to analyze data seamlessly and make a powerful impact through your projects.
Table of Contents (13 chapters)

Binary classification with logistic regression

Instead of attempting to predict what the total first-day return will be, we are going to attempt to predict whether the IPO will be one we should buy for a trade or not. It is here that we should point out that this is not investment advice and is for illustrative purposes only. Please don't run out and start day trading IPOs with this model willy-nilly. It will end badly.

Now, to predict a binary outcome (that's a 1 or 0/yes or no), we will start with a model called logistic regression. Logistic regression is actually a binary classification model rather than regression. But it does utilize the typical form of a linear regression; it just does so within a logistic function.

A typical single variable regression model takes the following form:

Here, t is a linear function of a single explanatory variable, x. This can, of...