Artificial intelligence
Surrogate Modeling
Understanding aerodynamics of airfoils typically requires high-fidelity computational fluid dynamics simulations, which can be resource intensive. Surrogate models allow us to arrive at the outcome with considerably less work.
IC Component logo identification
Circuit board components form a significant portion of the electronics industry. Based on hand-labeled data, this project uses computer vision to locate and identify component-specific logos with 95% accuracy.
Airline delays
Flight delays are a nuisance for travelers worldwide. Based on 2015 data, this project examines data across six main airlines, characterizing airline delay frequency and predicting delay types for each airline.
HYperlocal fog prediction
Localized fog is a problem for terrestrial sensor networks. This project predicts hyper-localized fog presence based on past and present weather conditions. Comparing different ML prediction models, an 93% accuracy is obtained.
modern tools
for relevatn applications