Improving short-term prandial blood glucose outcomes for people with type 1 diabetes, a complex disease that affects nearly 10 million people worldwide. We aim to leverage semi-supervised learning to identify unlabelled meals in time-series blood glucose data, develop meal-scoring functions, and explore causal machine-learning techniques.
A model instantiation function that can take a variety of parameter settings.
This function should serve as a hyperparameter training loop: loads the specified dataset, the train_test split, the model choice, the model hyperparameter settings, and perform the fit, test predict, and error calculation.
The function should also write the training runs training logs to models/training_logs directory.
A model loading function that can load a model from 0_meal_identification/meal_identification/models
The model should be loadable either fitting or predicting new data
A model storing function that stores the trained models in 0_meal_identification/meal_identification/models
Training Pipeline Script
Code Location: /bg_control/0_meal_identification/meal_identification/meal_identification/modeling/train.py
Requirements:
Useful Resources: