In this project, we implement a softmax classifier and a K-nearest-neighbor algorithm from scratch and train them. We do not use any DL library, only classic math libraries are required (numpy, math, mathplotlib...).
To understand how the project was built, follow the notebooks in the following order:
- knn_nosol.ipynb
- softmax_nosol.ipynb
All these notebooks call files in the nndl and utils folder. The requirements for this project are listed in requirements.docx .
These notebooks originate from uncompleted assignements from Pr. Kao's class : Neural Networks and Deep Learning at UCLA. I implemented all the missing parts.