- Author: Tanay Agrawal
- ISBN: 1484265785
- Year: 2020
- Pages: 185
- Language: English
- File size: 3.3 MB
- File format: PDF, ePub
- Category: Python
Dive into hyperparameter tuning of machine learning models and concentrate on what hyperparameters are and how they work. This publication discusses different methods of hyperparameters tuning, from the basics to advanced methods.
This really is a step-by-step guide to hyperparameter optimization, beginning with what hyperparameters are and how they affect different facets of machine learning models. It then goes through some fundamental (brute force) algorithms of hyperparameter optimization. Further, the writer addresses the problem of time and memory constraints, using dispersed optimization methods. Next, you’ll discuss Bayesian optimization for hyperparameter hunt, which learns from the previous history.
The publication discusses different frameworks, for example, Hyperopt and Optuna, which implement sequential model-based global optimization (SMBO) algorithms. During these talks, you’ll concentrate on different facets like the production of search spaces and the spread optimization of those libraries.
Hyperparameter Optimization in Machine Learning generates an understanding of how these algorithms work and how you can use them in real-life information science problems. The final chapter summarizes the role of hyperparameter optimization in automatic machine learning and finishes with a tutorial to make your own AutoML script.
Hyperparameter optimization is a tedious task, so sit back and let these algorithms do your work. What You Will Learn
- Discover how modifications in hyperparameters affect the model’s functionality.
- Apply Various hyperparameter tuning algorithms to information science issues
- Utilize Bayesian optimization methods to create an efficient machine learning and deep learning versions
- Distribute hyperparameter optimization using a bunch of machines
- Approach automatic machine learning using hyperparameter optimization
Who This Book Is For
Pros and students working together with machine learning.