watcher-ai-samples

Datasets, Java code and Spark notebook of Watcher-AI.

Structure

• spark-notebook-resources: the spark notebook file

• watcher-trainer: a Scala project to train and compare ML models

• watcher-ai-samples: a Java application wich uses the trained ML models to predict values

To start this notebook in your environment, please follow this steps:

• Step 1 : Download this version of Spark Notebook

• Step 2 : Save the folder "spark-notebook-resources" inside the folder where Spark are installed

• Step 3 : Go to the Terminal and inside the folder of Spark and run: ./bin/spark-notebook

• Step 4 : Go to the folder "spark-mllib-sample/spark-notebook-resources/" and open the file called "watcher-ai-core-ml"

All of this code and samples are available to download and test your own ML algorithm with Apache Spark MLlib.

This content is part of the presentation: Machine learning in practice with Spark MLlib: An intelligent data analyzer, presented at Strata + Hadoop World Singapore 2016

Authors

 Flavio Clésio - [email protected] / [email protected]
 J.P. Eiti Kimura  - [email protected]