DanielKerrigan/vis4ml-class-notebooks

Visualization for machine learning notebooks

Visualization for Machine Learning Notebooks

Contents

• 00_vis_intro.ipynb: An introdcution to visualization with Altair.
• 01_altair_continued.ipynb: Answers to questions about Altair.
• 02_data_exploration.ipynb and 02_data_exploration_solutions.ipynb: Histograms, scatter plot matrices, and confusion matrices.
• 03_projections.ipynb and 03_projections_solutions.ipynb: Histograms, scatter plot matrices, and confusion matrices.
• 04_altair_continued.ipynb: Answers to questions about Altair relating to interaction.
• 05_pdp_ice.ipynb and 05_pdp_ice_solutions.ipynb: PDP and ICE plots.
• 06_shap.ipynb and 06_shap_solutions.ipynb: Shapley values.
• 07_neural_networks_part_1.ipynb and 07_neural_networks_part_1_solutions.ipynb: Activation embeddings and saliency maps.
• 08_neural_networks_part_2.ipynb and 08_neural_networks_part_2_solutions.ipynb: Activation matrix.

Installation

# clone the repo
git clone https://github.com/DanielKerrigan/vis4ml-class-notebooks.git
cd vis4ml-class-notebooks

# create a virtual environment
conda create -n vis4ml python=3.10
conda activate vis4ml

# install dependencies
pip install -r requirements.txt

# run the server
jupyter notebook

You can also run these notebooks on Google Colab.