🤖 Grasp Stability Prediction
Deep model that predicts whether a robotic grasp is stable using tactile sensor images, deployed as an interactive web app.
PyTorch
ResNet18
MLP
Gradio
Featured Projects
Three projects that best represent how I think, build, and ship.
🪙 Euro Coin Classifier (ViT)
Fine-tuned a Vision Transformer to classify Euro coin denominations from images, achieving perfect test accuracy on a custom dataset.
ViT
PyTorch
Transformers
Computer Vision
🗺️ Asymmetric TSP Solver (ATSP-MTZ)
Exact solver for the Asymmetric Traveling Salesperson Problem using a Mixed Integer Programming model with MTZ subtour elimination.
C++
MIP
Gurobi
Graphs
More Projects
Additional work in tabular ML, dashboards, detection, and optimization.
🍷 Wine Quality Prediction
Predicts wine quality from physicochemical features with an end-to-end pipeline including preprocessing, feature selection, and model comparison.
Python
scikit-learn
Pandas
🔧 Steel Plate Fault Detection Dashboard
Streamlit dashboard comparing unsupervised and supervised methods for fault classification on steel plates, with interactive visualizations.
Streamlit
Python
UCI Dataset
🍽️ Multiple Object Recognition (Plates)
Detects repeated circular objects (plates) using gradient-based keypoints, DBSCAN clustering, and ellipse fitting, deployed as a Gradio app.
Python
OpenCV
DBSCAN
Gradio
🚚 Multi-Traveler Optimization (mTSP-MTZ)
Extended TSP to multiple agents, formulating a MIP model that partitions cities across salespeople while minimizing total cost.
C++
MIP
Gurobi
🌐 Minimum Spanning Tree (MST)
Implemented Kruskal’s and Prim’s algorithms to build minimum spanning trees connecting all nodes with the lowest possible cost.
C++
Graphs
Algorithms
💎 Prize-Collecting Steiner Tree (PCST)
Built exact PCST models to select high-value nodes and connections by trading off collected prizes against edge costs.
C++
MIP
Graph Optimization