Data Science Solutions on Azure

Understand and learn the skills needed to use modern tools in Microsoft Azure. This book discusses how to practically apply these tools in the industry, and help drive the transformation of organizations into a knowledge and data-driven entity. It provides an end-to-end understanding of data science life cycle and the techniques to efficiently productionize workloads.

The book starts with an introduction to data science and discusses the statistical techniques data scientists should know. You'll then move on to machine learning in Azure where you will review the basics of data preparation and engineering, along with Azure ML service and automated machine learning. You'll also explore Azure Databricks and learn how to deploy, create and manage the same. In the final chapters you'll go through machine learning operations in Azure followed by the practical implementation of artificial intelligence through machine learning.

Data Science Solutions on Azure will reveal how the different Azure services work together using real life scenarios and how-to-build solutions in a single comprehensive cloud ecosystem.

WHAT YOU'LL LEARN

* Understand big data analytics with Spark in Azure Databricks
* Integrate with Azure services like Azure Machine Learning and Azure Synaps
* Deploy, publish and monitor your data science workloads with MLOps
* Review data abstraction, model management and versioning with GitHub

WHO THIS BOOK IS FOR

Data Scientists looking to deploy end-to-end solutions on Azure with latest tools and techniques.

JULIAN SOH is a cloud solutions architect with Microsoft, focusing in the areas of artificial intelligence, cognitive services, and advanced analytics. Prior to his current role, Julian worked extensively in major public cloud initiatives, such as SaaS (Microsoft Office 365), IaaS/PaaS (Microsoft Azure), and hybrid private-public cloud implementations.

PRIYANSHI SINGH is a data scientist by training and a data enthusiast by nature specializing in machine learning techniques applied to predictive analytics, computer vision and natural language processing. She holds a master’s degree in Data Science from New York University and is currently a Cloud Solution Architect at Microsoft helping the public sector to transform citizen services with Artificial Intelligence. She also leads a meetup community based out of New York to help educate public sector employees via hands on labs and discussions. Apart from her passion for learning new technologies and innovating with AI, she is a sports enthusiast, a great badminton player and enjoys playing Billiards. Find her on LinkedIn at https://www.linkedin.com/in/priyanshi-singh5/

PART I - INTRODUCTION TO DATA SCIENCE AND ITS RISE TO PROMINENCE

Chapter 1 Data Science in the modern enterprise

What is Data Science

The Data Scientists' tools and lingo

Ethics and ethical AI

Significance of Data Science in organizations

Case Studies of applied Data Science

Chapter 2 Most important Statistical Tehniques in Data Science

Top Statistical Tehniques Data Scientists need to know

Supervised Learning

Unsupervised Learning

Regression/Classification/ Forecasting

Bayesian method

Time series analysis

Linear regression

Sampling methods

Reinforcement Learning

PART 2 - MACHINE LEARNING IN MICROSOFT AZURE

Chapter 3 Basics of data preparation and data engineering

Ingesting disparate data sources

Preparing data for analysis

Data Exploration

Feature Engineering

Chapter 4 Introducing Azure Machine Learning

AzureML- DataStores/ Datasets

Azure ML Compute/CLusters/inference/batch-realtime

Azure ML Service- Training and Building

Azure ML Service- Deploying

Azure ML Service- Pipelines

Azure ML Studio/Designer

Azure Automated Machine Learning (AutoML)

Hyperparameter Training

Azure ML- Security

Case Study

PART 3 - AZURE DATABRICKS

Chapter 5 Spark and Big Data

Spark and Hadoop

What is Big Data?

Why Spark is the platform of choice for Big Data

Challenges with Big Data

Chapter 6 Azure Databricks Basics

What is Azure Databricks

Azure Databricks from the Data Engineers' perspective

Azure Databricks from the Data Scientists' perspective

Chapter 7 Azure Databricks

Deploying the Azure Databricks workspace

Creating and Managing Clusters

Creating and managing users and groups

Managing Databricks Notebooks

Using Databricks Notebooks

DBFS

Connecting to ADLS

Sample Notebook(s)

PART 4 - OPERATIONALIZING DATA SCIENCE

Chapter 8 Machine Learning Operations

Operationalization concepts and DevOps

MLOps in Azure

MLFlow in Azure Databricks

Git

Chapter 9 Practical ML

Introducing use cases in the different industries

Democratizing AI through ML