/feature-selection-for-reinforcement-learning

Primary LanguagePython

Feature-selection-for-reinforcement-learning

Given:

MDP_policy.py Feature_Table.xlsx MDP_Original_data.csv MDP_Original_data

Goal: Maximize ECR and IS values using RL

To Evaluate:

  1. Run "python MDP_policy.py -input Training_data_ECR.csv" for Best ECR
  2. Run "python MDP_policy.py -input Training_data_IS.csv" for Best IS

To Execute:

  1. Keep all the data files in the same folder
  2. Run "python MDP_function.py"
  3. Run "python MDP_policy.py -input Training_data.csv" for Best ECR with Training_data.csv generated from 2
  4. Run "python MDP_function2.py"
  5. Run "python MDP_policy.py -input Training_data.csv" for Best IS with Training_data.csv generated from 4

Note: "python MDP_function2.py" automatically saves two files as best-ecr and isv based on the calculations that were carried out. Step 5 has to done to check if the value generated is actually the best ECR/ISV.