Shorten Link: http://bit.ly/eccb2021
ECCB web: https://www.iscb.org/ismbeccb2021-program/tutorials#tut5
Day 1: https://docs.google.com/presentation/d/1ohJcikVyI6VLi3ermEm2vZIjzP7mK2aK1j145JSS7xQ/edit?usp=sharing
Day 2: https://docs.google.com/presentation/d/1-Pp8KCVX5fpIM7OR1WYhr1gZLPhGsY6OYZGZkDU-f6A/edit?usp=sharing
Last year materials: https://github.com/ML-Bioinfo-CEITEC/ECCB2020
Day 1: https://www.youtube.com/watch?v=a8sp6xpBJ7g&t=25s&ab_channel=ISCB
Day 2: https://www.youtube.com/watch?v=PuSVakewtF4&t=22s&ab_channel=ISCB
- 01_One_Neuron.ipynb
- 02_MNIST.ipynb
- 03_CNN - conv. neural network on images
- 04_Fine-tuning (ResNet)
- 05_CNN on G4, DNA one-hot encoding
- 06_Regularization - dropout
- 07_TFjs web page
- 08_RNN classification
- 09_RNN generation
- 10_Integrated_Gradients
- 11_Grad-CAM
- Practice, or use a version with data preprocessing
https://ml-bioinfo-ceitec.github.io/ECCB2021/
Where to go next:
- (perfect for beginners eager to use ML in biology) MIT course Computational Systems Biology: Deep Learning in the Life Sciences https://mit6874.github.io/
- (slightly more advanced) Yannic Kilcher’s channel explaining various DL papers: https://www.youtube.com/channel/UCZHmQk67mSJgfCCTn7xBfew
- (theoretical) Grad-CAM - slightly more matematical https://glassboxmedicine.com/2020/05/29/grad-cam-visual-explanations-from-deep-networks/
- (theoretical) Overview of Explainability for Digital Pathology (a little bit longer): https://is.muni.cz/th/etz46/krajnansky-diploma-explainability.pdf
- see Day2 slides 68&69 in out presentation