Book Image

Mastering Machine Learning on AWS

By : Dr. Saket S.R. Mengle, Maximo Gurmendez
Book Image

Mastering Machine Learning on AWS

By: Dr. Saket S.R. Mengle, Maximo Gurmendez

Overview of this book

Amazon Web Services (AWS) is constantly driving new innovations that empower data scientists to explore a variety of machine learning (ML) cloud services. This book is your comprehensive reference for learning and implementing advanced ML algorithms in AWS cloud. As you go through the chapters, you’ll gain insights into how these algorithms can be trained, tuned, and deployed in AWS using Apache Spark on Elastic Map Reduce (EMR), SageMaker, and TensorFlow. While you focus on algorithms such as XGBoost, linear models, factorization machines, and deep nets, the book will also provide you with an overview of AWS as well as detailed practical applications that will help you solve real-world problems. Every application includes a series of companion notebooks with all the necessary code to run on AWS. In the next few chapters, you will learn to use SageMaker and EMR Notebooks to perform a range of tasks, right from smart analytics and predictive modeling through to sentiment analysis. By the end of this book, you will be equipped with the skills you need to effectively handle machine learning projects and implement and evaluate algorithms on AWS.
Table of Contents (24 chapters)
Free Chapter
1
Section 1: Machine Learning on AWS
3
Section 2: Implementing Machine Learning Algorithms at Scale on AWS
9
Section 3: Deep Learning
13
Section 4: Integrating Ready-Made AWS Machine Learning Services
17
Section 5: Optimizing and Deploying Models through AWS
Appendix: Getting Started with AWS

The importance of model optimization

Very few algorithms produce optimized models at the first attempt. This is because the algorithms might need some parameter tuning from the data scientist to improve their accuracy or performance. For example, the learning rate for deep neural networks that we mentioned in Chapter 7, Implementing Deep Learning Algorithms, needs to be manually tuned. A low learning rate may lead the algorithm to take longer (and hence be more expensive if we're running it on the cloud), whereas a high learning rate might miss the optimal set of weights. Likewise, a tree with more levels may take more time to train, but could create a model with better predictive capabilities (although it could also cause the tree to overfit). These parameters that direct the learning of the algorithms are called hyperparameters, and contrary to the model parameters (for...