Machine Learning for Daylight Prediction
Patrick Duhirwe
Machine Learning Model Development: Developing and training CNN, GRU, and CNN + GRU models for predicting indoor illuminance.
Model Performance Optimization and Evaluation: Tuning model parameters and evaluating performance using metrics like R2, RMSE, and MAE.
Generalization and Sensor Grouping: Testing models' generalization on unseen data and grouping illuminance sensors for analysis.
Computational Efficiency Analysis: Comparing models in terms of training time, prediction speed, and model size.
Like this project
Posted Feb 23, 2024
This project uses CNN and GRU models to predict indoor lighting from daylight, aiming for energy efficiency and enhanced environmental design.
Likes
0
Views
5
