Hello! I'm Gus, an AI Engineer at with a deep passion for mathematics, physics, machine learning, AI, and problem-solving. I have a B.S. in physics from the University of Massachusetts Amherst and a master's in Data Science from Harvard. Professionally, I've transitioned from my role as an ML Research Engineer at AIMdyn Inc., where I developed Koopman-based machine learning models for autonomous system learning, to my current position as an AI Engineer at Lowe's. Here, I apply my expertise in AI and machine learning to drive innovation and efficiency. Beyond my professional work, I actively participate in data science competitions on Kaggle and run a YouTube channel, 'Math & Physics Fun with Gus', where I share insights on complex mathematical theories and data science concepts. With my background in physics and data science, I'm constantly seeking new challenges to apply my skills in solving complex problems and advancing the field of AI and machine learning.
Old GitHub GusLovesMath
-
Local LLM Training on Apple Silicon - This repository contains the resources and documentation for the project "Local LLM Training on Apple Silicon", where the Llama3 model was fine-tuned to efficiently solve verbose mathematical word problems on an Apple Silicon device with 16 GPUs.
-
NeuralForecast | Nixtla NeuralForecast - Features comprehensive neural forecasting models including NBEATS, enhancing stock market analysis, prediction accuracy, and model training time.
-
High-Accuracy Brain Tumor Classification using CNN - A CNN developed with TensorFlow 2.16 and GPU acceleration achieves a 99.7% accuracy rate in classifying brain tumors, incorporating advanced techniques like data augmentation and adaptive learning rate adjustments.
-
Stock Forecasting with Multi-Step Stacked LSTM - Utilizes stacked LSTM models for detailed Tesla stock price forecasting, with extensive documentation on LSTM network theory and practical application.
-
Ozone (O3) AQI Trends in High-Impact U.S. Regions - Applies SARIMAX and Holt-Winters models to analyze and predict ground-level ozone fluctuations, contributing to better environmental policy making.
