Model-Card-Project-IFT6390

Model card creation for Linear classifier using CrowdFlower AI gender predictor dataset. This is an exploratory SVM linear classifier which predicts the gender of the user based on its Tweets and profile description.

Prerequisites:

1. Python 3 and later
2. Sklearn
3. Numpy
4. Matplotlib
5. Preprocessor
6. Pylatex
7. Latex
8. Pdflatex

Running the model:

In terminal, execute the following commands:

1- Fetch the dataset

$ python run.py fetch

2- Train the model

$ python run.py train

3- Evaluate the model

$ python run.py evaluate

4- Generate the model card and the report

$ python run.py build_paper

Versioning:

Version 1.0

Authors:

Houda Saadaoui, Vijaya Lakshmi Kuruba

License:

This project is licensed under the MIT License

This model was developed as a requirement of the course IFT 6390 at the University of Montreal in Winter 2021

• This is an exploratory SVM linear classifier which predicts the gender of the user based on its Tweets and profile description.
• SKlearn API is utilised for model creation. We have used SGD classifier with Hinge loss with L2 penality.
• Released under MIT License

For any questions about the model, contact houda.saadaoui@umontreal.ca, vijaya.lakshmi.kuruba@unomtreal.ca