Official implementation of "Dynamic Label Injection for Imbalanced Industrial Defect Segmentation" accepted at VISION Workshop - ECCV 2024
-
Updated
Jun 5, 2025 - Python
Official implementation of "Dynamic Label Injection for Imbalanced Industrial Defect Segmentation" accepted at VISION Workshop - ECCV 2024
Python notebook describing the entire process of developing a prediction model including: Exploratory Data Analysis, Data Preprocessing, and Model Development
This project uses an enhanced WGAN with GCN layers to tackle class imbalance in fraud detection. By converting data into graph structures, it generates realistic synthetic samples, improving balance and accuracy compared to traditional methods like SMOTE.
Machine learning project predicting bank term deposit subscriptions using customer demographics and campaign data. Features multiple ML models (Random Forest, Logistic Regression, Decision Tree), SMOTE for class balancing, and interactive Streamlit app for real-time predictions.
Cyclist safety risk assessment application for Austin using Random Forest classification with SMOTE for class balancing. Interactive Streamlit web app predicts cycling incident severity based on time, location, and environmental conditions to promote safer cycling practices.
Use PySpark to predict the success of a terrorist attack using different machine learning approaches
Análisis de Sentimientos en Reseñas de Películas: Proyecto que utiliza NLP (tokenización, lematización, eliminación de contracciones y detección de idioma) para clasificar reseñas según 8 emociones distintas, implementando un modelo de análisis y visualizando patrones con matplotlib.
💰 Python CLI pipeline for loan default classification. Includes data cleaning, visualization (Matplotlib), balancing, Decision Tree (Sklearn), and interactive prediction browser
Add a description, image, and links to the class-balancing topic page so that developers can more easily learn about it.
To associate your repository with the class-balancing topic, visit your repo's landing page and select "manage topics."