fau-masters-collected-works-cgarbin
Goto Github
PK
Name: Christian Garbin CS master's and Ph.D. collected works
Type: Organization
Bio: Work created during FAU's computer science master's and Ph.D. (data science, machine learning, ...)
Location: United States of America
Blog: https://cgarbin.github.io/about/
Christian Garbin CS master's and Ph.D. collected works's Projects
Experiments with LLM agents using Microsoft's AutoGen
CAP5615 Intro to Neural Networks class at FAU, Summer 2018
Data science with Pandas and NumPy: EDA, binning, distribution functions, simulations, regression analysis
Computer vision using traditional classifiers, neural networks, and transfer learning
CAP6619 Deep Learning FAU CS master's Fall 2018
Information retrieval from basic concepts (tokenization, stop word removal, stemming, TF-IDF, etc.)
A datasheet for the ChestX-ray8 dataset, a.k.a. ChestX-ray14
A model card for the CheXNet model.
A preprocessor and visualizer for CheXpert
COT-5930 Image processing using MATLAB
COT6405 Analysis of algorithms Spring 2020
Federated learning: literature review and experiments with the Google sample code
COT6930 Natural Language Processing, Spring 2019
Exploring data visualization with Facets and Streamlit
Template for datasheet for datasets
What is "accuracy"? The effect of changing the decision threshold on a model's accuracy.
Dropout vs. batch normalization: effect on accuracy, training and inference times - code for the paper
A "chat with your data" example: using a large language models (LLM) to interact with our own (local) data. Everything is local: the embedding model, the LLM, the vector database. This is an example of retrieval-augmented generation (RAG): we find relevant sections from our documents and pass it to the LLM as part of the prompt (see pics).
IEEE ICMLA 2019 Data Science Tutorial - using data to answer questions
A tool to compare multiple large language models (LLMs) side by side
Summarizing with LLMs: Using an LLM to understand GitHub issues without reading each post in detail.
Are machines "learning" anything? This repository explores some of the concepts from the book "Artificial Intelligence, a guide for thinking humans", by Melanie Mitchell.
Template for model cards
Methodically "unobfuscate" the winner of the 1988 International Obfuscated C Contest entry, the incredible code that prints the "twelve days of Christmas" song.
Ridge, elastic net, and logistic regressions implemented without using any statistical or machine learning library. All steps are done by hand, using matrix operations as much as possible.
Exploring SHAP feature attribution for image classification
Writing good Jupyter notebooks: logically organized, clearly documented decisions and assumptions, easy-to-understand code, flexible (easy to modify) code, resilient (hard to break) code