This is a workshop prepared for the Machine Learning class of the "INF8215- AI: Methods and Algorithms" class given in Polytechnique Montreal in Fall 2018.
The main file LAB3.ipynb gives the instructions for the whole lab. The workshop is divided in three parts.
In the first part, students have to implement from scratch using python and only numpy a Softmax Classifier following the instructions given in the file LAB3.ipynb . A template is given in the file SoftmaxClassifier.py in order to make the task easier for students.
In the second part, the students are given the "Shelter Animal Outcomes" dataset (taken from the kaggle website https://www.kaggle.com/c/shelter-animal-outcomes) and have to perform data preprocessing on it using scikit-learn and pandas. The goal is to make the data usable for Machine Learning models training. The students need to justify their choices for every single attribute.
In the third part the students must train a few Machine Learning models using the preprocessed data and scikit-learn along with the model implemented in the first part. Then they have to improve the performance of one selected model.
The solution can be found in the folder solution .