You are on page 1of 3

Assignment No.

Markov Decision Process

G-23
Assignment No. 8

Title: Markov Decision Process

Problem Definition and Aim of Experiment: To study and apply Markov


process for modelling manufacturing processes.

Objective of Experiment
1. To understand Reinforcement Learning
2. To understand Markov Decision Process (MDP) used to formulate the
reinforcement learning problems
3. To analyse the problem and draw Markov’s Chain
4. To find the probability of sequence of steps for given number of steps.
5. To find the probability of sequence when initial and final states are
given.

Theory

1. What is Reinforcement Learning? (Schematic, Explanation)


2. How Reinforcement Learning works?
3. Reinforcement Learning Terminology
4. Reinforcement Learning Work Flow
5. Characteristics of Reinforced Learning
6. How RL is Different from Other Machine Learning Paradigms?
7. Reinforced Learning Classification
 Positive Learning
 Negative Learning
8. Reinforced Learning Categorical Algorithms Classification
 Value based
 Policy based
 Actor critic
 Model free
 Model based
9. Markov Decision Process
10. Bellman Equation

Problem Statement
Write problem statement of already available one type of learning
example from available resources. The data must be in CSV format. The
aim is to extract features from given dataset and establish training data.

Procedure
Write and follow the standard procedure for your selected type of learning
example from available resources assigned to you.

Program Code
Write Algorithm, draw flow-chart of the Problem Statement selected by /
assigned to you. Also enclose a print-out of your Python Program Listing
(*.txt).

G-24
NOTE: No hand-written Program Listing is expected; it should be a Print-out!

Program Input
Also attach a print-out of the Dataset in form of *.csv file (comma
separated values) as an input.

Program Output
 Attach a print-out of your program output (result) and Conclusion.

G-25

You might also like