Mousteph

Descent

This project aims to display any function R -> R or R^2 -> R and to compare different gradient descent methods to find the minimum of this function in a visual way.

PortFin

PortFin is a project designed to compare the performance of dynamically allocated investment portfolios with the market, S&P500. It leverages historical data of the assets of the S&P500 to generate new portfolios each year and allocates assets accordingly. All the information is then compiled into a PDF report.

Object_detection_GPGPU

An object detector for processing videos, delivering quasi real-time results using CUDA.

Anomaly_detection

Anomaly Detection using Machine Learning Models on the UGR'16 Dataset. Explore the effectiveness of Isolation Forest, One-Class SVM, and XGBoost in identifying anomalies in a subsampled dataset from July 2016.

MLOPS_Wine_Quality

Predicting Red Wine Quality with MLOPS: A complete system for wine quality prediction, data management, and data drift monitoring. This project uses Docker containers, Streamlit for user interfaces, PostgresSQL to save the data, and Kafka for data communication, providing a robust pipeline for wine quality predictions.

Skin_disease

This project is designed for classifying various skin diseases using the HAM10000 dataset. It leverages a trained model, explains predictions using LIME, and provides multiple interfaces for users, including a server, a graphical user interface, a command-line interface, and an API.

FTML_Project

PYBD_project