You are previewing Microsoft Azure Machine Learning.
O'Reilly logo
Microsoft Azure Machine Learning

Book Description

Explore predictive analytics using step-by-step tutorials and build models to make prediction in a jiffy with a few mouse clicks

In Detail

This book provides you with the skills necessary to get started with Azure Machine Learning to build predictive models as quickly as possible, in a very intuitive way, whether you are completely new to predictive analysis or an existing practitioner.

The book starts by exploring ML Studio, the browser-based development environment, and explores the first step—data exploration and visualization. You will then build different predictive models using both supervised and unsupervised algorithms, including a simple recommender system. The focus then shifts to learning how to deploy a model to production and publishing it as an API.

The book ends with a couple of case studies using all the concepts and skills you have learned throughout the book to solve real-world problems.

What You Will Learn

  • Learn to use Azure Machine Learning Studio to visualize and pre-process data

  • Build models and make predictions using data classification, regression, and clustering algorithms

  • Build a basic recommender system

  • Deploy your predictive solution as a Web service API

  • Integrate R and Python code in your model built with ML Studio

  • Explore with more than one case study

  • Downloading the example code for this book. You can download the example code files for all Packt books you have purchased from your account at http://www.PacktPub.com. If you purchased this book elsewhere, you can visit http://www.PacktPub.com/support and register to have the files e-mailed directly to you.

    Table of Contents

    1. Microsoft Azure Machine Learning
      1. Table of Contents
      2. Microsoft Azure Machine Learning
      3. Credits
      4. About the Author
      5. Acknowledgments
      6. About the Reviewers
      7. www.PacktPub.com
        1. Support files, eBooks, discount offers, and more
          1. Why subscribe?
          2. Free access for Packt account holders
          3. Instant updates on new Packt books
      8. Preface
        1. What this book covers
        2. What you need for this book
        3. Who this book is for
        4. Conventions
        5. Reader feedback
        6. Customer support
          1. Downloading the color images of this book
          2. Errata
          3. Piracy
          4. Questions
      9. 1. Introduction
        1. Introduction to predictive analytics
          1. Problem definition and scoping
          2. Data collection
          3. Data exploration and preparation
          4. Model development
          5. Model deployment
        2. Machine learning
          1. Types of machine learning problems
            1. Classification
            2. Regression
            3. Clustering
          2. Common machine learning techniques/algorithms
            1. Linear regression
            2. Logistic regression
            3. Decision tree-based ensemble models
            4. Neural networks and deep learning
        3. Introduction to Azure Machine Learning
          1. ML Studio
        4. Summary
      10. 2. ML Studio Inside Out
        1. Introduction to ML Studio
        2. Getting started with Microsoft Azure
          1. Microsoft account and subscription
          2. Creating and managing ML workspaces
        3. Inside ML Studio
          1. Experiments
            1. Creating and editing an experiment
            2. Running an experiment
            3. Creating and running an experiment – do it yourself
        4. Workspace as a collaborative environment
        5. Summary
      11. 3. Data Exploration and Visualization
        1. The basic concepts
          1. The mean
          2. The median
          3. Standard deviation and variance
          4. Understanding a histogram
          5. The box and whiskers plot
            1. The outliers
          6. A scatter plot
        2. Data exploration in ML Studio
          1. Visualizing an automobile price dataset
            1. A histogram
            2. The box and whiskers plot
            3. Comparing features
            4. A snapshot
        3. Do it yourself
        4. Summary
      12. 4. Getting Data in and out of ML Studio
        1. Getting data in ML Studio
          1. Uploading data from a PC
          2. The Enter Data module
          3. The Data Reader module
            1. Getting data from the Web
              1. Fetching a public dataset – do it yourself
            2. Getting data from Azure
        2. Data format conversion
        3. Getting data from ML Studio
          1. Saving a dataset on a PC
            1. Saving results in ML Studio
            2. The Writer module
        4. Summary
      13. 5. Data Preparation
        1. Data manipulation
          1. Clean Missing Data
          2. Removing duplicate rows
          3. Project columns
          4. The Metadata Editor module
          5. The Add Columns module
          6. The Add Rows module
          7. The Join module
        2. Splitting data
        3. Do it yourself
        4. The Apply SQL Transformation module
        5. Advanced data preprocessing
          1. Removing outliers
          2. Data normalization
          3. The Apply Math Operation module
          4. Feature selection
            1. The Filter Based Feature Selection module
            2. The Fisher Linear Discriminant Analysis module
          5. Data preparation beyond ready-made modules
        6. Summary
      14. 6. Regression Models
        1. Understanding regression algorithms
        2. Train, score, and evaluate
          1. The test and train dataset
          2. Evaluating
            1. The mean absolute error
            2. The root mean squared error
            3. The relative absolute error
            4. The relative squared error
            5. The coefficient of determination
        3. Linear regression
        4. Optimizing parameters for a learner – the sweep parameters module
          1. The decision forest regression
        5. The train neural network regression – do it yourself
        6. Comparing models with the evaluate model
          1. Comparing models – the neural network and boosted decision tree
        7. Other regression algorithms
        8. No free lunch
        9. Summary
      15. 7. Classification Models
        1. Understanding classification
          1. Evaluation metrics
            1. True positive
            2. False positive
            3. True negative
            4. False negative
            5. Accuracy
            6. Precision
            7. Recall
            8. The F1 score
            9. Threshold
            10. Understanding ROC and AUC
            11. Motivation for the matrix to consider
        2. Training, scoring, and evaluating modules
        3. Classifying diabetes or not
          1. Two-class bayes point machine
          2. Two-class neural network with parameter sweeping
        4. Predicting adult income with decision-tree-based models
        5. Do it yourself – comparing models to choose the best
        6. Multiclass classification
          1. Evaluation metrics – multiclass classification
        7. Multiclass classification with the Iris dataset
          1. Multiclass decision forest
          2. Comparing models – multiclass decision forest and logistic regression
        8. Multiclass classification with the Wine dataset
          1. Multiclass neural network with parameter sweep
          2. Do it yourself – multiclass decision jungle
        9. Summary
      16. 8. Clustering
        1. Understanding the K-means clustering algorithm
        2. Creating a K-means clustering model using ML Studio
        3. Do it yourself
        4. Clustering versus classification
        5. Summary
      17. 9. A Recommender System
        1. The Matchbox recommender
          1. Types of recommendations
          2. Understanding the recommender modules
            1. The Train Matchbox recommender
              1. The number of traits
              2. The number of recommendation algorithm iterations
            2. The Score Matchbox recommender
            3. The evaluate recommender
        2. Building a recommendation system
        3. Summary
      18. 10. Extensibility with R and Python
        1. Introduction to R
        2. Introduction to Python
        3. Why should you extend through R/Python code?
        4. Extending experiments using the Python language
          1. Understanding the Execute Python Script module
            1. Creating visualizations using Python
          2. A simple time series analysis with the Python script
          3. Importing the existing Python code
          4. Do it yourself – Python
        5. Extending experiments using the R language
          1. Understanding the Execute R Script module
          2. A simple time series analysis with the R script
          3. Importing an existing R code
            1. Including an R package
          4. Understanding the Create R Model module
          5. Do it yourself – R
        6. Summary
      19. 11. Publishing a Model as a Web Service
        1. Preparing an experiment to be published
          1. Saving a trained model
        2. Creating a scoring experiment
        3. Specifying the input and output of the web service
        4. Publishing a model as a web service
          1. Visually testing a web service
        5. Consuming a published web service
        6. Web service configuration
        7. Updating the web service
        8. Summary
      20. 12. Case Study Exercise I
        1. Problem definition and scope
        2. The dataset
        3. Data exploration and preparation
          1. Feature selection
        4. Model development
        5. Model deployment
        6. Summary
      21. 13. Case Study Exercise II
        1. Problem definition and scope
        2. The dataset
        3. Data exploration and preparation
        4. Model development
        5. Model deployment
        6. Summary
      22. Index