Machine learning is a newly-reinvigorated field. It promises to foster many technological advances that may improve the quality of our lives significantly, from the use of the latest, popular, high-gear gadgets such as smartphones, home devices, TVs, game consoles and even self-driving cars, and so on. Of course, for all of us in the circles of high education, academic research and various industrial fields, it offers more challenges and more opportunities.
Whether you are a CS student taking a machine learning class or a scientist or an engineer entering the field of machine learning, this text helps you get up to speed with machine learning quickly and systematically. By adopting a quantitative approach, you will be able to grasp many of the machine learning core concepts, algorithms, models, methodologies, strategies and best practices within a minimal amount of time. Throughout the text, you will be provided with proper textual explanations and graphical exhibitions augmented not only with relevant mathematics for its rigor, conciseness, and necessity but also with high-quality examples.
The text encourages you to take a hands-on approach while grasping all rigorous, necessary mathematical underpinnings behind various ML models. Specifically, this text helps you:
- Understand what problems machine learning can help solve
- Understand various machine learning models, with the strengths and limitations of each model
- Understand how various major machine learning algorithms work behind the scene so that you would be able to optimize, tune, and size various models more effectively and efficiently
- Understand a few state-of-the-art neural network architectures such as Convolutional Neural Networks (CNNs), Recurrent Neural Networks (RNNs), and Autoencoders (AEs), and so on
- More importantly, you can learn how to train and run practically usable deep learning models on macOS and Linux-based instances with GPUs
Solutions to exercises are also provided to help you self-check your self-paced learning.
HENRY H. LIU, PHD, is a computer software performance practitioner and a machine learning researcher with a physicist background. During his prior physicist career, he achieved high-impact results with extraordinarily accurate theoretical research and predictive modeling on the motion of particles traveling at nearly the speed of light. After jumped to computers, he applied his research and predictive modeling skills to computer software system performance challenges and achieved amazingly accurate forecasts & predictions in special event driven, unusually high traffic production environment. He is interested in leveraging his knowledge in advanced mathematics and extensive research and practicing experience in performance to help advance machine learning for solving real application problems.
