Held tutoring and review sessions for Data Structures and Program Development course taught in C++ and Introduction to Computer Science course taught in Python.
Worked on a proof-of-concept prototype of an offline cost-effective system capable of predicting roadway concentrations of ultrafine particles (UFP) using a low-power camera and a combination of sensors.
•
Created an algorithm to count the number of cars in each frame in Python using openCV.
•
Preprocessed and condensed different time series parameters such as car count, wind speed, and particulate matter concentrations into a unified time series using pandas.
•
Built a recurrent neural network (RNN) using the unified time series data to predict the UFP concentration on roads using PyTorch.
•
Conducted data exploration and experimentation to model the UFP concentrations as multiple decaying functions using pandas and NumPy.
