/teaching/pgm/Recent content in Probabilistic Graphical Models on girish labHugo -- gohugo.ioen/teaching/pgm/proj/Mon, 01 Jan 0001 00:00:00 +0000/teaching/pgm/proj/ <ol> <li>Implement a program, which</li> <li>Which on input the MN DAG (as adjacency matrix), and samples from the joint/and/or marginal distribution, learns the joint distribution.</li> <li>Which on input the MN DAG and the joint distribution, and a conditional probability query, outputs the probability The project should be made into a public github repo, with proper readme files.</li> </ol> <ul> <li> <p>Bayes Ball</p> <ol> <li>Which on input X,Y,Z, checks conditional independence using d-seperation. All examples on the textbook need to be given as test cases.</li> <li>Linear Time Bayes Ball algorithm and analysis with latex scribe.</li> </ol> </li> <li> <p>Prove theorems regarding equivalence of Markov Networks. Equivalence of CI statement and d-seperability.</p> </li> <li> <p>IMNet : <a href="https://pdfs.semanticscholar.org/6314/2e3f05bfc1e1394a0f46af341837a40fe3b8.pdf">https://pdfs.semanticscholar.org/6314/2e3f05bfc1e1394a0f46af341837a40fe3b8.pdf</a></p> </li> <li> <p>Ancestral Graph Markov Model: <a href="https://projecteuclid.org/download/pdf_1/euclid.aos/1031689015">https://projecteuclid.org/download/pdf_1/euclid.aos/1031689015</a></p> </li> <li> <p>Causality and do calculus (Pearl[220])</p> </li> <li> <p>Expert Systems in Medical Field: <a href="https://projecteuclid.org/euclid.ss/1177010888">https://projecteuclid.org/euclid.ss/1177010888</a></p> </li> </ul>