Summary

Example of how to use data and task

License

CC-BY-SA 3.0

Tags

api demo

Feature Processing

No preprocessing or scaling of features are done

Parameters

SVM C=1.0 with Gaussian kernel width=1.0

Operating System

agnostic

Code

See source code, clientapi/api_demo.sh clientapi/mlprocess.py

Software Packages

mldata-utils available at: http://mloss.org/software/view/262/

Description

Demonstrate how to use the application programming interface (API) of mldata.org.

The code detects the task type {binary classification, multiclass, regression} and creates the appropriate SVM.

Location of labels vs. examples, as well as the training and test splits are all taken automatically from the task file.

URLs

http://mloss.org/software/view/262/

Publications

revision 1

by cong on 2010-11-30 14:19

No one has posted any comments yet. Perhaps you would like to be the first?

Leave a comment

This item was downloaded 0 times and viewed 6012 times.

Method mldata api demo has been applied the following tasks:

Submitter	Task (version)	Curve	Score	Date
cong	pima binclass (1)	-	0.817164179104  predictions	2010-11-30 14:20
cong	pima binclass roc (3)		0.791336570406  predictions	2010-11-30 14:22
cong	iris multiclass (1)	-	1.0  predictions	2010-11-30 14:22
cong	pima binclass (1)	-	0.817164179104  predictions	2013-06-27 11:19
cong	boston housing scaled regression (1)	-	0.396992035548  predictions	2010-12-02 18:00
cong	boston housing regression (1)	-	8.56191478601  predictions	2010-12-02 18:01