Parallel streaming decision trees

author: Yossi Richter, IBM Haifa Research Lab

Description

A new algorithm for building decision tree classifiers is proposed. The algorithm is executed in a distributed environment and is especially designed for classifying large datasets and streaming data. It is empirically shown to be as accurate as standard decision tree classifiers, while being scalable to infinite streaming data and multiple processors.

You might be experiencing some problems with Your Video player.

*Slides*
0:00	Parallel streaming decision trees
0:17	Why decision trees?
0:50	Previous work
2:56	Streaming parallel decision tree
4:23	Iterative parallel decision tree
5:10	Building an on-line histogram
6:10	Merging two histograms
6:44	Example of the histogram
7:21	Pruning
9:04	IBM Parallel Machine Learning toolbox
9:50	Results: Comparing single node solvers
9:55	IBM Parallel Machine Learning toolbox
10:49	Results: Comparing single node solvers
11:27	Results: Pruning
11:56	Speedup (Strong scalability)
13:08	Weak scalability
13:40	Algorithm complexity
14:05	Summary
14:53	Thank You
16:21	- Questions

Lecture rating

People found this lecture:
Worth seeing
because it is:
Valuable and informative
Well presented
Easily understandable
Acceptably recorded
You need to login to cast your vote.

Report a problem or upload files

If you have found a problem with this lecture or would like to send us extra material, articles, exercises, etc., please use our ticket system to describe your request and upload the data.
Enter your e-mail into the 'Cc' field, and we will keep you updated with your request's status.