Bandits simulator lab01

About this Multi-Armed Bandits problem:

Multi-Armed Bandits is a classical game where there exists an Agent with K different actions which can be activated by choice, one at a time. When an action is taken, the Agent interacts with the environment by that particular action. The environment then will produce a signal as the reward for the corresponding actions, this reward may follow a random distribution.
The Agent repeats this process for T rounds, each round one action.
The goal is to maximize the total sum of the receiving rewards.

How to run simulator:

⚠️ Your computer is required to have jdk (jre + javac), without whom the simulator can not be executed.

After assuring your requirements, run in your terminal: bash run.sh

For those who use window operating system, search keywords "compile java command line window" for more detail.

Another ultility you can use is this online coding website, make sure to include run.sh file so that you can execute the program with bash run.sh

How to complete:

finish: Agent_method.java

given:
- the total number of actions by Construtor method: nAction
- the reward of the previous action taken: pastReward, this value is 0 when this function is called the first time
- the id of the previous action taken: pastActionId, this value is -1 when this function is called the first time
the body of the class is yours to design, create table of values, create method to compute information returns...
return the id of the action you choose

Notice: Consider both exploration and exploitation for the best performance

After Simulation:

The program is coded to produce a trace file, namely result.csv where all the information during the simulation is printed.

Also, at the end of the simulation, you are expected to see your own score at the bottom of the terminal output.

🏁 For more competitive results and figures, you can submit file Agent_method.java upto this website where your codes will be tested with multiple scripts, your scores are ranked with others.

🏁 About submittion onto ranking website: Make sure your solution is legid, as once you submit your solution, its score is cummulatively computed. Meaning, your score is the average score of all submittions.

General warning:

😏 feel free to create new classes to support your code