predictive analytics: getting started with amazon machine learning

©2015, Amazon Web Services, Inc. or its affiliates. All rights reserved

Getting Started with Machine Learning

Guy Ernest, BDM MLgernest@amazon.com

Main Takeaways

• Machine Learning is a focus in Amazon• ML is big and growing• ML is easy and will be used by everyone

How to be successful in Business

E*BI

RTML

EC2ECSElastic Beanstalk

RedshiftEMR

KinesisElasticSearch

Amazon MLSpark ML

What do your kids learn in Math Class

4 Steps to Solving Math Problems

• Posing the right question• Real world to computation formulation• Computation• Computation formulation to the real world

4 Steps to Solving Math ML Problems

• Posing the right question• Real world to computation formulation• Computation• Computation formulation to the real world

=Data

=Application

=Business Problem

The circle of ML

Application

E*

Data

Model Customer

Front end team

Data Engineering team

Analysts / DS team

DevOps team

And a few more examples…

Fraud detection Detecting fraudulent transactions, filtering spam emails, flagging suspicious reviews, …

Personalization Recommending content, predictive content loading, improving user experience, …

Targeted marketing Matching customers and offers, choosing marketing campaigns, cross-selling and up-selling, …

Content classification Categorizing documents, matching hiring managers and resumes, …

Churn prediction Finding customers who are likely to stop using the service, upgrade targeting, …

Customer support Predictive routing of customer emails, social media listening, …

What do you need to know to use Machine Learning?

ML Model is a function to split Space

Historical Data Model Building Prediction

What is my color?

And what is mine?

Why more data is better?

Less Data More Data Even More Data

Why more attributes are better?

Less Attributes More Attributes Even More Attributes

Where to Split?

Data

Engineering

Why Clean Data is better?

Messy Data Cleaner Data Fantasy Data

Gray Area

Recall and Precision

• Which mistake do you prefer?

Linear Regression

Parameters ComplexityLow complexitySparse Data

High complexityDense Data

Neural Networks

Speech Recognition

Face recognition

OCRThe Face Neuron

Machine Translation

こんにちはשלוםEvent/Doc

Classification

Light ML Amazon ML Statistical ML Deep Learning

Images

Input

Natural Language

Your Fortune Cookie

“Stop writing heuristic code, and start building predictive models”