S04 - A Soft-Labeled Self-Training Approach

[kerautret]

General info

• [x] Conference: ICPR 2016
• [x] Source paper: paper S04
• [x] Papers title: A Soft-Labeled Self-Training Approach
• [x] GitHub repo: https://github.com/AlexanderMey/Soft_Labeled_Self_Learning
• [x] Submission date: 1/12/2016
• [x] Platform: Ubuntu 17.10
• [x] Language: Matlab R2017b
• [x] Results to be reproduced: tab 2, 3, 4, 5, 6

• [x] Reproduction main step instructions:

   Give the main steps to reproduce the figure/tab
   Preparation:
   - Unzip Soft_Labeled_Self_Learning.zip
  
   Tab. 2:
   1. Load `Nearest Mean/Banana_Experiment_NM.mat`
   2. Run `NearestMeanSS.m`
  
   Tab. 3:
   1. Move `Nearest Mean/Gauss2_LS.mat` to `Nearest Mean/Gauss2_NM.mat`
   2. Load `Nearest Mean/All_Gauss_NM.mat`
   3. Run `NearestMeanSS.m`
  
   Tab. 4:
   1. Load `Least Squared/All_Gauss_LS.mat`
   2. Run `LeastSquaresSS.m`
  
   Tab. 5:
   1. Load `Nearest Mean/All_Sets_NM.mat`
   2. Run `NearestMeanSS.m`
  
   Tab. 6:
   1. Load `Least Squared/All_Sets_LS.mat`
   2. Run `LeastSquaresSS.m`

Reviewer feedback

• [x] Tested platform: Ubuntu 17.10
• [x] Result reproduced: Matlab R2017b
• [x] Reviewer comment:
  • Matrix Nearest Mean/Gauss2_NM.mat doesn't exist: second line of Tab. 3 cannot be tested, tests are executed with Nearest Mean/Gauss2_LS.mat
    • Tab 4: doesn't give always results after execution
    • Some results can differ

Details Results

• Execution without errors:
  • [ ] without changes
  • [x] with small changes
  • [ ] with a lot of changes
• Results are obtained:
  • [ ] always
  • [x] sometimes
  • [ ] never
• Results correspond to the paper's figures/tables:
  • [ ] exactly
  • [ ] with negligible differences
  • [x] with differences
  • [ ] with a lot of differences