Step 1: run startup.m to add dependencies
Step 2: run demo/small_20plot.m to see an interesting toy example.
Step 3: run demo/demo_digits_sample.m to see the most sophisticated usage 
which include taking a subsample, building a knn graph, and running all
active learning strategies for 2 repetitions (EER has been commented out). 
Step 4: modify demo_digits_sample.m to create your own problem. (ref 
  

  

    

      sigma_opt_codes/demo/small_20plot.m
    
    

         Line 13
      in
      0b62ff1
    
  
  

    


        

          

          
 load DBLP20_v1 
        
    
  


  

  

    

      sigma_opt_codes/demo/demo_digits_sample.m
    
    

         Line 2
      in
      0b62ff1
    
  
  

    


        

          

          
 load digits_scikit.csv; 
        
    
  


)

Also included:
data/ some other graphs I used. Their sources are detailed in the paper.

Contact: 
Yifei Ma. [email protected]