Garbage Classification
Machine learning system that classifies garbage into 6 classes to support environmental awareness and waste reduction.
Highlights
- Built and deployed a garbage classification ML model (6 classes) to support environmental awareness and waste reduction
- Compared VGG16, VGG19, ResNet50, and ResNet152; selected ResNet50 for deployment (~92% accuracy, ~0.015 loss)
Objective
The main objective was to build a garbage classification system. This helps reduce pollution and raise community awareness of environmental protection
Build ML models for classifying 6 classes: metal, glass, organic, paper, battery, plastic
Explored deploying a Machine Learning application into a production environment
Studied and compared the performance of different image classification models
Result
Compared the performance of 4 models: VGG16, VGG19, ResNet50, and ResNet152
Selected ResNet50 for deployment due to its balance between accuracy and hardware efficiency
Achieved ~92% accuracy and ~0.015 loss.
Key tasks: Data Cleaning | EDA | Data Processing | Build Model (VGG16)
Skills/Tools: Python | Machine Learning | API
Outcome: Kaggle Notebook (Garbage Classification)