moonGen
Classifying difficulty of climbing routes on the MoonBoard apparatus using novel machine learning / deep learning techniques
This is the project repository for the CS230 Winter 2020 course project
Running Experiments
For all experiments, make sure to change the root directory indicated in each script!
To run baseline experiments:
- First, generate data for baseline experiments by running
\scripts\baseline\gen_baseline_data.py - Batch-run baseline experiments by running
\scripts\baseline\run_baseline_models.py
To run neural network experiments, simply execute \scripts\pytorch\run_pytorch_models.py (after making sure that the root directories are proper)
Current Status (DONE)
Completely finished:
- Scraping
- Baseline PyTorch framework: Dense (Fully-Connected), GCN
- Baseline statistical learning models
- Evaluation metrics
- Batch-run of all experiments
Potential Future Items
- Convolutional neural network as another baseline (work off of problem-hold matrix)
- Autoencoder for hold embedding extraction