The assignments for AIST4010 Foundation of Applied Deep Learning, which are Kaggle competitions on image, sequence and graph tasks.

Table 1. Topic and competition result of different tasks

• To install package, type the following command in a terminal:

pip install <package_name>

1. numpy: scientific computations
2. pandas: importing and exporting data from / to csv
3. matplotlib: graph plotting
4. sklearn: machine learning algorithms, data preprocessing, evaluation metrics
5. natsort: sorting of list data
6. torch: deep learning related algorithms

7. torchvision: dataset and data augmentation for images
8. facenet_pytorch: pre-trained Inception Resnet v1 model on face dataset

9. Bio: loading protein sequence data
10. transformers: pretrained BERT model on protein dataset and trainer

11. torch_geometric: graph neural network algorithms

1. aist4010-asm<N>.ipynb

   • Code of the method used in Task N

2. AIST4010 Assignment <N>.pdf

   • Specification of Task N

1. output.csv

   • Output file of the code

2. aist4010-spring2022-a<N>-publicleaderboard.csv

   • The public leaderboard used to select the best model for Task N

3. aist4010-spring2022-a<N>-privateleaderboard.csv

   • The private leaderboard used to determine the rank of Task N

• 2nd place (out of 28 teams) in Task 1
• 1st place (out of 31 teams) in Task 2
• 2nd place (out of 8 teams) in Task 3

© 2022 Andes Kei