Seth Chatterton

Master of Business Analytics Student at MIT

Seth Chatterton

About Me

I am a current student in the Master of Business Analytics program at MIT. I have experience in data science, machine learning, optimization, and management consulting. My current research with Professor Dimitris Bertsimas is on using large language models for document understanding and structured information extraction. Previously, I was a Consultant at Sage Analysis Group, where I used computer vision, natural language processing, and simulation modeling to solve complex client problems. I'm interested in Data Scientist positions and similar roles, especially in roles where I can have direct business impact.

Resume

Email: setcha@mit.edu

LinkedIn: seth-chatterton

GitHub: setcha

Research Advisor: Dimitris Bertsimas

Projects

Optimal Routing for Electric Vehicle Last-Mile Delivery

12/2023

Vehicle Paths

Developed an optimal routing algorithm for electric vehicle last-mile delivery using Gurobi and Julia. The project focused on minimizing delivery time and maximizing efficiency, while respecting battery range and recharging time constraints.

Read the full report

Technologies: Gurobi, Julia

Language Modeling using Tree-Based Methods

12/2023

Language Modeling Tree

Implemented tree-based models for natural language processing tasks. The project involved extensive data analysis and model training using Python libraries.

Read the full report

Technologies: Python, scikit-learn, pandas

Predicting Views of Viral Youtube Videos

12/2023

Viral Video Correlations

Created a predictive model to estimate the views of YouTube videos. The model was trained on various features such as video category, time published, colors in the thumbnail, and attributes of the title.

Read the full report

Technologies: Python, scikit-learn, pandas

Particle Swarm Optimization with Flocking and Genetic Programming

8/2018

Implemented a particle swarm optimization algorithm using the flocking behavior of birds for improved information diffusion, and genetic programming to choose flocking parameters and behaviors. The project aimed to create an algorithm capable of solving complex optimization problems.

View the poster

Technologies: Java, Python

U.S. Patent No. 10,118,670 - Strain Gauge Device and Approach

8/2016

Involved in the development of a novel strain gauge device for measuring the rigging tension on sailing vessels. The project resulted in a successful patent filing.

Read the patent

Technologies: Arduino, 3D printing