Ross Kimme

Ross
Kimme

About

I'm a student at UIUC exploring the intersection of data, systems, and problem-solving. I'm drawn to questions that combine quantitative thinking with real-world applications—whether that's understanding how policies shape behavior, optimizing physical systems, or building tools that help people make better decisions.

Projects

CMAPSS LSTM — Remaining Useful Life (In Progress)

LSTM neural network predicting Remaining Useful Life of jet turbine engines using NASA's CMAPSS benchmark. Sliding 30-cycle windows over 21 sensors feed a stacked LSTM for regression against true RUL labels. Generalizes to any safety-critical system with time-series sensor monitoring.

Deep Learning PyTorch Time Series

UIUC Course Explorer & GenEd Planner

Full-stack app aggregating GPA data, live section status, and professor ratings for UIUC course planning. Features a backtracking set-cover algorithm for optimal gen-ed combinations, prerequisite tree visualization, and grade distribution histograms.

Full-Stack Algorithms Data

ICE Deportation Risk Classifier

Deportation risk classifier and detention time predictor built on real ICE FOIA data. Designed to interpret black-box detention decisions and help nonprofits allocate legal resources.

ML Python FOIA Data

2024 Election Polling Analysis

Analyzed 600+ polls using confidence intervals and sampling distributions to identify discrepancies between polling predictions and actual 2024 election results.

Statistics Python Data

Skills

Python
NumPy
Pandas
SciPy
Scikit-learn
SQL
Matplotlib
Seaborn
Technical Writing
Jupyter Notebooks
Data Analysis and Statistics
Claude / Codex (IDE)

Contact