Automatic Machine Learning (AutoML) – Hands-on session

Agenda

• AutoML overview and what packages are available

• Hands-on session

1. Bad Loans (Binary classification)

2. Predict house prices (Regression)

– We will program using H20 Flow UI and Python API

– Call AutoML module from the above

Hands-on session Pre-requisites

H20 installation

1. Install Java 1.8+

2. https://h2o-release.s3.amazonaws.com/h2o/rel-weierstrass/6/index.html

Follow the instructions given in the “Download and Run” tab.

Pandas, Jupiter, and Python Binding for H20 installation

3. Follow the instructions in “Install in Python” Tab

Download Data Sets used in the session

4. Bad Loans DataSets

https://raw.githubusercontent.com/h2oai/app-consumer-loan/master/data/loan.csv

5. House Price Dataset:

 https://www.kaggle.com/c/house-prices-advanced-regression-techniques/data

Speaker: Bismayan Chakrabarti

Data Scientist, airisDATA Inc.