Time Series Classification

This project was developed for the course of Artificial Neural Networks and Deep Learning for the MSc. in Mathematical Engineering at Politecnico di Milano, A.Y. 2022/2023.

Description

.
├── README.md
├── misc
│   ├── report_material
│   ├── submission_template
│   │   ├── metadata
│   │   ├── model.py
│   │   ├── model_aug.py
│   │   └── scaler.gz
│   └── training_histories
├── notebook
│   ├── final_model.ipynb
│   ├── utils
│   │   ├── augmentation.py
│   │   ├── datasets.py
│   │   ├── dtw.py
│   │   ├── helper.py
│   │   ├── input_data.py
│   │   ├── models.py
│   │   ├── nemenyi.py
│   │   └── prototype_selection.py
│   ├── x_train.npy
│   └── y_train.npy
├── report
│   ├── bibliography.bib
│   ├── report.pdf
│   └── report.tex
└── requirements.txt

7 directories, 49 files

• final_model.ipynb is the main file that performs training of all the models and produces all the output images.
• report_material and training_histories are folders containing additional images, namely the plots of the different training histories of the various models.
• submission_template contains the code used to submit the model in the competition.
• x_train.npy and y_train.npy are the given dataset.
• utils contains the code to perform data augmentation.
• requirements.txt contains the necessary packages for the environment.

Authors

• Paolo Botta (@ploki99)
• Teo Bucci (@teobucci)
• Silvia Caresana (@silviacaresana)

Output

Check out the final report.pdf.

License

GNU GPLv3