Probabilistic-Graphical-Models

All homeworks of the Probabilistic Graphical Models class from ENS Paris-Saclay

Subjects

• DM1/Exercises_18oct2017.pdf
• DM2/Exercises_10nov2017.pdf
• DM3/HWK3_2017.pdf

Getting Started

• clone the repository on your computer
• open an Anaconda command prompt
• browse to the working directory:

cd path/to/your/directory/DM1/code

• launch the jupyter notebook:

jupyter notebook MVA_DM1_BUSA.ipynb

• run each cells using shift + enter

Code

The code for all assignments is available under the folder 'code'. All the implementations are done in python. To execute the code you should have jupyter installed in your computer as well as the necessary packages.

Explanations

The final reports are named MVA_DMX_BUSA_Victor.pdf where X = 1, 2 or 3 The latex version of the report is also available under the name: MVA_DMX_BUSA_Victor.tex with X = 1, 2 or 3

Notes:

• In homework 2 the answer to the question 3.b in Part 1 is no:
  • "Is the statement true in general?" the answer is no and you should provide a counter-example
• In homework 3 in the code (iPython Notebook):
  • There is a bug in my Viterbi algorithm implementation. I should get the same assignment using Viterbi algorithm and marginal distribution.