deep-learning.yaml

# 2-half-days deep learning intro class for science & engineering.
title: Intro to Deep Learning for Subsurface
environment: geoml
conda:
  - fastchan::fastai  # Installs PyTorch
pip:
  - skorch
curriculum:
  1a:
    - Quick course overview
    - Introductions
    - Deep learning vs shallow learning
    - Intro_to_NumPy_and_PyTorch.ipynb  # Done. Could add more PyTorch.
  1b:
    - What is a neural network?
    - Intro_to_neural_networks.ipynb  # Done.
    - Check out and feedback
  2a:
    - Check in and review
    - Intro_to_neural_networks.ipynb  # Continued.
  2b:
    - Image_classification_with_skorch.ipynb  # Done.
    - Requests_from_geofignet_API.ipynb  # Done.
    - Check out and feedback
extras:
  - All_the_activation_functions.ipynb
  - Image_classification.ipynb
  - Image_classification_with_PyTorch.ipynb
  - Intro_to_NumPy_for_ML.ipynb
  - Neural_networks_from_scratch.ipynb
  - Pandas_for_ML_data_management.ipynb
  - Pandas_for_timeseries.ipynb
  - Read_and_write_LAS.ipynb
  - Read_SEG-Y_with_segyio.ipynb
  - Reproducing_Roeth_and_Tarantola_1994.ipynb  # If we have time.
  - Transfer_learning_RESNET_inference_only.ipynb
  - Transfer_learning_RESNET_with_PyTorch_on_Colab.ipynb  # If time.
  - What_is_gradient_descent.ipynb  # Could add PyTorch.optim version of example.
scripts:
  - mlutils.py  # Copied to all notebook directories.
references:
  - Choromanska_etal.pdf
  - Dramsch_2020.pdf
  - Karpatne_etal_2017.pdf
  - ML_project_review_checklist.pdf
  - Raschka_2018.pdf
  - Reichstein_etal_2019.pdf