Artificial Intelligence

#Priyanka Mulagandla (pm3392)

Markov Process Solver using value and policy iterations:

Help with executions

usage: MdpSolver [-v] [-df DF] [-min MIN] [-tol TOL] [-iter ITER] [-input input_file]

Markov process solver

positional arguments: input_file Path to the input file

optional arguments: -v Enable verbosity for program runs -df DF discount factor in the range [0, 1]. Defaults to 1.0 -min MIN minimize values as costs, defaults to False which maximizes values as rewards -tol TOL tolerance. Defaults to 0.01 -iter ITER Maximum number of value iteration. Defaults to 100 -input input_file Takes the path given as input

Executing the programs

1. Unzip the mdpFinal.zip folder.

2. Compile the mdpFinal folder using the command line(in the folder and unzipped) javac -classpath "*:." *.java

3. The program can be executed with different options as follows:

Run the solver with default options

java -classpath "*:." MdpSolver -input ../lab3Tests/lab3_input6.txt

Run the solver with default options in verbose mode

java -classpath "*:." MdpSolver -v -input ../lab3Tests/lab3_input6.txt

Run the solver with a tolerance of 0.001

java -classpath "*:." MdpSolver -tol 0.001 -input ../lab3Tests/lab3_input6.txt

Run the solver with cost minimizer option

java -classpath "*:." MdpSolver -min true -input ../lab3Tests/lab3_input6.txt

Run the solver with a discount factor of 0.9

java -classpath "*:." MdpSolver -df 0.9 -input ../lab3Tests/lab3_input6.txt

NOTE: The input_file value given in the examples above are just for reference and have to be modified w.r.t proper test file paths