Pinned Repositories
CPET
Continuous Psychophysics with Eye Tracking: A MATLAB Based Program for Estimation of Contrast Sensitivity and Assessing Refractive Errors.
Multi-Task-House-Prediction
The goal of this project is to build a multi-task learning model using PyTorch Lightning that predicts both house prices (a regression task) and house categories (a classification task). This involves using a shared bottom neural network architecture with task-specific top layers, applying advanced machine learning techniques.
Patient-Readmission-Risk-Prediction-with-MLP-and-KAN
This project focuses on predicting patient readmission risk using two distinct neural network architectures: the traditional Multi-Layer Perceptron (MLP) and the novel Kolmogorov-Arnold Network (KAN). The project aims to compare the performance of these architectures in terms of accuracy and interpretability on a medical dataset.
Pandemic-Burnout-Predictor-ML-Insights-on-Employee-Well-being
An ML project leveraging causal inference to predict employee burnout rates during the pandemic, considering factors like work from home. Includes data analysis on mental health's impact on productivity.
Instagram-Engagement-Enhancer
This project leverages the Google Vision API and Latent Dirichlet Allocation (LDA) for topic modeling to analyze Instagram images. By evaluating the association between image content and user engagement, we provide data-driven recommendations to help influencers increase interaction on their posts.
Text-Analytics-Salary-Prediction
This project aims to build and test classification models to predict high and low salaries based on the text contained in job descriptions. The dataset used for this assignment is the "Job Salary Prediction" dataset from Kaggle.
MTL-Blog-Chatbot
The MTL Blog Chatbot project is an innovative solution designed to interactively provide users with information and updates from the MTL Blog website. Leveraging the power of the LlamaCpp Large Language Model (LLM), this chatbot offers responses across various categories such as news, eat/drink, things-to-do, and more.
Multi-class-Text-Classification-with-Transformers
This repository presents a detailed exploration and implementation of Transformer models for the task of multi-class text classification. Leveraging the extensive 20 Newsgroups dataset, this project showcases the end-to-end process of applying advanced NLP techniques to categorize text into one of 20 distinct categories.
Shakespeare-Text-RNN
This project aims to implement a Recurrent Neural Network (RNN) to generate text that mimics the style of Shakespeare's writings. We utilize PyTorch Lightning to streamline our implementation, focusing on data preprocessing, model building, training, and text generation.
COMP-551-Applied-Machine-Learning
In this project, we were tasked to implement a Naive Bayes model from scratch as well as a Bidirectional Encoder Representations from Transformers (BERT) with pretrained weights and to compare the results from training the IMDb review dataset.
ethanpirso's Repositories
ethanpirso/Multi-class-Text-Classification-with-Transformers
This repository presents a detailed exploration and implementation of Transformer models for the task of multi-class text classification. Leveraging the extensive 20 Newsgroups dataset, this project showcases the end-to-end process of applying advanced NLP techniques to categorize text into one of 20 distinct categories.
ethanpirso/MTL-Blog-Chatbot
The MTL Blog Chatbot project is an innovative solution designed to interactively provide users with information and updates from the MTL Blog website. Leveraging the power of the LlamaCpp Large Language Model (LLM), this chatbot offers responses across various categories such as news, eat/drink, things-to-do, and more.
ethanpirso/Shakespeare-Text-RNN
This project aims to implement a Recurrent Neural Network (RNN) to generate text that mimics the style of Shakespeare's writings. We utilize PyTorch Lightning to streamline our implementation, focusing on data preprocessing, model building, training, and text generation.
ethanpirso/pykan-cuda
Kolmogorov Arnold Networks (with cuda support)
ethanpirso/EthanNet
EthanNet is a deep convolutional neural network designed for image classification tasks, specifically benchmarked on the CIFAR-10 dataset. The model integrates VGG-like blocks and ResNet-like bottleneck blocks to achieve effective feature extraction while maintaining a balance between depth and computational efficiency.
ethanpirso/Fly-On-Time
This project aims to address flight delays by predicting them during the ticket purchasing process. The FlyOnTime demo uses historical data, weather conditions, and supervised learning models to forecast potential delays, offering a proactive solution compared to real-time tracking and post-event compensation services.
ethanpirso/Patient-Readmission-Risk-Prediction-with-MLP-and-KAN
This project focuses on predicting patient readmission risk using two distinct neural network architectures: the traditional Multi-Layer Perceptron (MLP) and the novel Kolmogorov-Arnold Network (KAN). The project aims to compare the performance of these architectures in terms of accuracy and interpretability on a medical dataset.
ethanpirso/Multi-Task-House-Prediction
The goal of this project is to build a multi-task learning model using PyTorch Lightning that predicts both house prices (a regression task) and house categories (a classification task). This involves using a shared bottom neural network architecture with task-specific top layers, applying advanced machine learning techniques.
ethanpirso/Instagram-Engagement-Enhancer
This project leverages the Google Vision API and Latent Dirichlet Allocation (LDA) for topic modeling to analyze Instagram images. By evaluating the association between image content and user engagement, we provide data-driven recommendations to help influencers increase interaction on their posts.
ethanpirso/MGSC-673-AI-and-Deep-Learning
Repository for Master of Management in Analytics course, Intro to AI and Deep Learning. The course covers topics in deep learning, architecture of neural networks, and practical implementation using PyTorch.
ethanpirso/Pandemic-Burnout-Predictor-ML-Insights-on-Employee-Well-being
An ML project leveraging causal inference to predict employee burnout rates during the pandemic, considering factors like work from home. Includes data analysis on mental health's impact on productivity.
ethanpirso/INSY-662-Data-Mining-and-Visualization
This project combines a Random Forest Classifier and a DBSCAN clustering model to predict Kickstarter campaign outcomes and categorize projects, offering insights into campaign success and project landscape.
ethanpirso/COMP-551-Applied-Machine-Learning
In this project, we were tasked to implement a Naive Bayes model from scratch as well as a Bidirectional Encoder Representations from Transformers (BERT) with pretrained weights and to compare the results from training the IMDb review dataset.
ethanpirso/Text-Analytics-Salary-Prediction
This project aims to build and test classification models to predict high and low salaries based on the text contained in job descriptions. The dataset used for this assignment is the "Job Salary Prediction" dataset from Kaggle.
ethanpirso/CPET
Continuous Psychophysics with Eye Tracking: A MATLAB Based Program for Estimation of Contrast Sensitivity and Assessing Refractive Errors.