Whispering Gallery Mode (WGM) study
Research Problem: predicting the value of environmental factors (temperature and pressure) from spectrum of light rays. In Physics, changing these environmental factors cause the crest of the spectrum to shift by varying amounts.
Research Task: (1) Implement non-parametric meta-learning algorithms (2) Train neural networks on our dataset (3) Evaluate their single-task and multi-task performance
Actions: (2000+ lines of code)
- Implemented non-parametric meta-learning algorithms
- Deep Neural Network (a baseline)
- Siamese Network algorithm
- Triplet Network algorithm
- Triplet Network algorithm & Temporal Convolutional Network (TCN)
- Implemented k-fold cross-validation to tune the hyperparameters of each model
- Trained and tested models on a Linux server, using tmux, a utility
- Experimented their performance on multi-task scenario in two ways:
- Meta-learning: Training on support dataset, then query dataset
- Multi-task learning: Training on a super-dataset with examples from both datasets
- Reported results to and discussed with my supervisor weekly
- Discussed with physicists to confirm and update expectations for research outcome
- Produced well-documented, modularized code, which I circulated with resident researchers who find them helpful on their projects
- Reported findings in lab meetings; received feedback and discussed possible applications of the algorithms on other projects
- Created an easy-to-use interface for all models in Jupyter notebook
Results:
- Reduced the mean-square-error loss from 0.016 to 0.000081
- Identified the limitation of algorithms used
- Compared to the baseline, increasingly complex non-parametric algorithms behave perform significantly better.
- In particular, hyperparameter settings that work well for one dataset also works well on others
- However, the amount of improvement from pre-training on supporting dataset diminishes with greater model complexity
- Training on mixed 'super-dataset' yeilds result similar to the baseline but is much worse than the single-task performance