Siddhardha Penmetsa

Lifelong science learner with a research background in Astrophysics. I enjoy solving problems using statistics and machine learning, primarily with Python’s data science libraries. Proficient in SQL and can adapt to R and Matlab as needed. Experienced in leveraging LLMs (like GPT-4) for workflow efficiency.

Deep Learning project to classify types of supernovae - Led planning and weekly meetings. Member in an End to End Deep learning group project at Waterloo. Collaborated via github. Implemented normalising flows(NF) algorithm to approximate supernova light curves and passed it through a Convolutional Neural Network that predicts types of supernovae. Developed visualisation modules for the project.

Analysis of the 21 cm hydrogen emission line: Modeled early universe physics to detect the 21 cm hydrogen emission line from the Epoch of Reionization, employing Python for simulation and analysis. Crafted Python scripts to generate artificial datasets incorporating the emission line feature, successfully identifying the Epoch of Reionization signals. Authored a comprehensive scientific report, complete with data visualizations, detailing the methodology, analysis, and detection of the Epoch of Reionization.