Colin Commans

I'm a student at UC San Diego majoring in Math and Computer Science.

At UCSD, I have embraced the opportunity to explore and refine my passion for problem solving by taking on rigorous and challenging courses. A lifelong love of puzzles has always fueled my interest in math, and I try to use coding as a tool to creatively visualize and ground the abstract concepts I have encountered in my studies. I certainly plan to keep learning math on my own after college and hope to continue solving complex problems creatively wherever my career ends up taking me. Please check out my projects below to learn more about my interests and skills, and feel free to reach out with any questions!

Email  /  Resume  /  Github  /  LinkedIn

Projects

I'm interested in machine learning algorithms, deep learning, and generative (particularly adversarial) AI.

Re-implemented the Least Squares Generative Adversarial Network (LSGAN) framework, exploring challenges such as vanishing gradients and mode collapse. Experimented with different parameterizations and loss functions (Hinge, L1) on datasets like EMNIST, KTH-TIPS, and Tiny ImageNet.

Developed a sentiment analysis model to predict positive or negative text from Amazon book reviews. Data preprocessing, model optimization via GridSearch, and a Flask web app enable real-time sentiment predictions.

Miscellanea

• Programming Languages: Java, C/C++, MATLAB, SQL, JavaScript, Python
• Developer Tools: Git, VS Code, Jupyter Notebook
• Libraries: Matplotlib, Numpy, Pandas, Scikit-learn, PyTorch
• Frameworks: JUnit, Flask
• MS Office: Word, Excel, PowerPoint

• Deep Learning
• Machine Learning
• Statistical Learning
• (Un)Constrained Optimization
• (Non)Linear Optimization
• Data Structures & Algorithms
• Probability & Statistics
• Differential Geometry
• Complex Analysis
• Real Analysis
• Abstract Algebra
• Topology

Website source code taken from Jon Barron's.