Название: Evolutionary Deep Learning: Genetic algorithms and neural networks (MEAP v12)
Автор: Micheal Lanham
Издательство: Manning Publications
Год: 2023
Страниц: 443
Язык: английский
Формат: pdf, epub
Размер: 23.3 MB

Discover one-of-a-kind AI strategies never before seen outside of academic papers! Learn how the principles of evolutionary computation overcome deep learning’s common pitfalls and deliver adaptable model upgrades without constant manual adjustment.

Evolutionary Deep Learning is a guide to improving your Deep Learning models with AutoML enhancements based on the principles of biological evolution. This exciting new approach utilizes lesser- known AI approaches to boost performance without hours of data annotation or model hyperparameter tuning.

Google Colab notebooks make it easy to experiment and play around with each exciting example. By the time you’ve finished reading Evolutionary Deep Learning, you’ll be ready to build deep learning models as self-sufficient systems you can efficiently adapt to changing requirements.

As we move into a new decade the Deep Learning technology is becoming the mainstream machine learning work horse for a wide variety of industries and organizations. While the technology has advanced quickly in the last decade many practitioners new or experienced often struggle with building function real-world networks to solve their problems. This book attempts to address some of those shortcomings by introducing the application of evolutionary algorithms for the optimization and betterment of DL systems.

More than a couple of decades ago I learned in tandem evolutionary algorithms and neural networks. At the time both branches of what was considered by some pseudo-sciences were still in their infancy. Both struggled to be relevant and applicable, but I fell in love with them because they embraced deep fascinations of mine, understanding how life came to be and how we think. In fact, many have been fascinated and attempted to understand how we humans have evolved to think like we do. A concept that is now the foundation for artificial intelligence but in the last few years have been separated in practice. Instead, we rely on more traditional scientific methods to try and develop human like intelligence.

While this book won’t go so far as open the door to discovering new forms of artificial intelligence, it does set the foundation to think about the way we do develop ML systems. Where instead of relying on our human bias and/or experience we introduce robust systems based on evolution to automatically optimize deep learning networks. Showing you several techniques, you can apply to move your deep learning networks beyond your expectations.

To enjoy this book, it is recommended you have some exposure working with and understanding how a Deep Learning framework like Keras or PyTorch function. While you don’t have to be an expert you should be comfortable enough identifying the key components of a Deep Learning system and how they function. If you understand at a high level the mathematical concepts that power deep learning you should also be fine learning new evolutionary methods. Since this is a hands-on book it is also recommended you have experience running Python code as a data scientist or Machine Learning engineer.