In this repo, I apply various statistical methods (Bayesian) and neural networks in Python to analyze and examine relevant topics within the field of computational linguistics. This was part of a 400-level course designed for us to engage in interdisciplinary methods to investigate linguistic phenomena.
Each problem set includes an .ipynb file that can be opened by Github, but can only run on Jupyter Notebook. The following is the topic of each problem set:
Problem Set 1:
1.1 - Neural Networks and Back Propagation
1.2 - Beta Distribution and Homophones
1.3 - Surprisal (Bayesian Methods)
Problem Set 2:
2.1 - Semantic Representations
2.2 - Semantic Networks
2.3 - Noun Classes in Tsez
Problem Set 3 (3.3 cancelled):
3.1 - Cross-linguistic Color Representations
3.2 - Bilingual Semantic Networks
GPA received: 4.0
For further questions, email me at josh.gelua@mail.utoronto.ca
Special thanks to Professor Suzanne Stevenson (Department of Computer Science, UofT) and Julia Watson (TA, Department of Computer Science) for curating this course. Credit for the supplementary provided functions goes to them.
Here's a link to Prof. Stevenson's website for her bio and publications: http://www.cs.toronto.edu/~suzanne/