Mastering Azure Machine Learning

More Information
  • Load and preprocess large datasets using Azure Data Factory
  • Train Machine Learning models using Azure Databricks
  • Learn to deploy machine learning models on Azure Kubernetes Services
  • Learn how to use DSVMs and Notebooks for Plotting and Embedding
  • Normalize data, and fill missing values using Spark in Azure Databricks
  • Implement feature extraction with word embedding using Natural Language Processing
  • Implement a distributed model training using Uber’s Horovod Estimator
  • Use the Catalyst optimizer to improve query performance in Spark
  • Train Machine learning model using Azure ML Compute and Azure Databricks
  • Explore how to track and optimize the model performance over time
  • Implement a Real-time Scoring Service on Azure Kubernetes Services for automated Machine Learning deployments

Data is eating the world and most data professionals need to make sense of it. The massive increase of data requires complex distributed systems, powerful algorithms and scalable cloud infrastructure in order to compute insights, train models and deploy them at scale.

This book is a comprehensive guide to build large end-to-end Machine Learning pipelines in the cloud using Azure and Machine Learning services. Starting with Azure Data Science VMs, Notebooks and Azure Machine Learning Service you will perform and schedule common data loading and preparation technique using Azure Databricks and Azure Data Factory. Next, you will cover NLP, classical Machine Learning techniques such as ensemble techniques, time-series forecasting as well as Deep Learning for classification and regression. Leveraging state-of-the-art technologies, you will learn how to train, optimize and tune models using Automated-ML and Hyperdrive. You will learn to perform distributed training using Azure ML Compute. Later, you will learn different monitoring and optimization techniques in order to measure training performance in Spark using Azure Databricks. Finally, you will learn to deploy models to Kubernetes using Azure Machine Learning Service

By the end of this book, you will master Azure Machine Learning Service and be able to build, optimize and operate scalable Machine Learning pipelines in Azure.

  • Implement highly-scalable end-to-end Machine Learning pipelines on Azure
  • Train and optimize advanced Deep Learning models on Spark using Azure Databricks
  • Deploy Machine Learning models for batch and real-time scoring using Azure Kubernetes Services
Page Count 490
Course Length 14 hours 42 minutes
ISBN 9781789807554
Date Of Publication 22 Nov 2019


Christoph Körner

Christoph currently works as a Big Data, Advanced Analytics, and Artificial Intelligence Consultant at Microsoft in Dublin where he helps his clients architecting, building and optimizing Big Data and Data Science platforms in the cloud. Previously, he was the Big Data Technical Lead at T-Mobile Austria where he designed, implemented and operated large scale data, analytics and prediction pipelines on Hadoop. He also authored the 3 books: Deep Learning in the Browser (for Bleeding Edge Press), Learning Responsive Data Visualization and Data Visualization with D3 and AngularJS (both for Packt).

Kaijisse Waaijer

Kaijisse Waaijer is an experienced technologist, specializing in Data Platforms, Machine learning, and IoT. Kaijisse currently works for Microsoft EMEA as a Data Platform Consultant, specializing in Data Science, Machine learning and Big Data. She constantly works with customers across multiple industries as their trusted tech advisor, helping them optimize their organizational data creating better outcomes and business insights that drive value, using Microsoft technologies. Her true passion lies within the Trading Systems Automation and applying deep learning and neural networks to achieve advanced levels of prediction and automation.