Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Multiclass ecoc #527

Merged
merged 5 commits into from May 16, 2012
Merged

Multiclass ecoc #527

merged 5 commits into from May 16, 2012

Conversation

pluskid
Copy link
Contributor

@pluskid pluskid commented May 15, 2012

A new Encoder and a python_modular script that automatically cover all encoder/decoder pairs. When the test labels are given, the accuracy is already computed. Here's an example output (only for some sense ...)

Testing with 6 encoders and 4 decoders


      encoder+decoder          codelen    time       accuracy  
 Discriminant+AED              9          1.690      0.4441    
 Discriminant+ED               9          1.770      0.8338    
 Discriminant+HD               9          1.750      0.6274    
 Discriminant+IHD              9          1.780      0.3779    
       Forest+AED              27         5.400      0.3224    
       Forest+ED               27         5.420      0.8139    
       Forest+HD               27         5.740      0.8281    
       Forest+IHD              27         5.650      0.3662    
          OVO+AED              45         2.710      0.4210    
          OVO+ED               45         2.690      0.7484    
          OVO+HD               45         2.700      0.9687    
          OVO+IHD              45         2.710      0.9676    
          OVR+AED              10         1.310      0.9452    
          OVR+ED               10         1.310      0.9452    
          OVR+HD               10         1.300      0.8911    
          OVR+IHD              10         1.320      0.8972    
  RandomDense+AED              23         13.870     0.9320    
  RandomDense+ED               23         14.030     0.9267    
  RandomDense+HD               23         14.510     0.8840    
  RandomDense+IHD              23         13.950     0.9096    
 RandomSparse+AED              35         6.440      0.3331    
 RandomSparse+ED               35         6.340      0.8705    
 RandomSparse+HD               35         9.390      0.9238    
 RandomSparse+IHD              35         7.510      0.9359    

CECOCForestEncoder::CECOCForestEncoder()
{
m_num_trees = 3;
m_parameters->add(&m_num_trees, "num_trees", "number of trees");
Copy link
Member

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

please use SG_ADD instead

Copy link
Contributor Author

Choose a reason for hiding this comment

The reason will be displayed to describe this comment to others. Learn more.

Thanks! I have fixed this.

sonney2k pushed a commit that referenced this pull request May 16, 2012
@sonney2k sonney2k merged commit a60c648 into shogun-toolbox:master May 16, 2012
sonney2k pushed a commit that referenced this pull request May 22, 2012
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Labels
None yet
Projects
None yet
Development

Successfully merging this pull request may close these issues.

None yet

2 participants