GiorgiaAuroraAdorni/CAT-optimal-hybrid-solver

The CAT Optimal Hybrid Solver is a tool designed to tackle the cross array task (CAT) activity designed to assess algorithmic thinking skills in the context of K-12 education.

CAT Optimal Hybrid Solver

Run Program

  cd ExactMethods/BreadthFirstSearch/
  clang++ -std=c++11 -stdlib=libc++ -g BfsAlgo.cpp -o bfs
  ./bfs NAME_GARPH

Description

The program takes a .txt file containing a cat array as input, prints the number of minimum moves to complete the problem and also prints the program to get to the result.

CAT_BFS* + GA

  cd MetaMethods/GA2/
  clang++ -std=c++11 -stdlib=libc++ -g GA2.cpp -o ga
  ./ga

Note: to change the graph, you must modify the main function in GA2.cpp.

CAT_BFS* + Hybrid approach

  cd PseudoML/MetaMethods/ExpAlgo2
  clang++ -std=c++11 -stdlib=libc++ -g ExpAlgo.cpp -o exA
  ./exA

Note: to change the graph, you must modify the main function in ExpAlgo.cpp.

Run Reinforcement Learning

  cd SlurmModel/
  cd chose the model you prefer

Look the run file to check fthe ormat of run some need a number to choose file and AI, and other need change run file.

Example:

  cd SlurmModel/Model_8_slurm1
  python3 run_Model.py 4

If you need training graph enter it in a folder of an AI and run

  python3 -m tensorboard.main --logdir logs --load_fast=false