Neural network and Decision tree experimentation

This project is a personnel experimentation that take a the data-set zoo to try to classify it using supervised learning.

• Python: 2.7
• Sklearn version: 0.18

Results:

For the Neural Network

Some results running the script for 600 iteration using one hidden layer:

Id	Neurone	Momentum	Learning rate	Error rate	mean-squared error
1	20	0.6	0.15	0.23	1.88
2	20	0.1	0.2	0.25	1.96
3	15	0.01	0.4	0.46	2.84
4	15	0.6	0.08	0.25	1.32
5	10	0.72	0.025	0.12	0.625
6	10	0.5	0.01	0.42	1.33
7	10	0.5	0.4	0.73	3.38
8	10	0.5	0.4	0.45	1.77
9	5	0.65	0.15	0.43	4.06
10	5	0.67	0.2	0.47	5.04

For the Decision Tree

You will find the results of the decision tree following this link Decision Tree Results