Machine Learning Projects

Convolutional Neural Networks

Project 1 : Fashion Image Classification

Classifys images based on type of apparel, footwear, accesory into 10 possible output neurons

• 60,000 training images , - 10,000 test images
• Data Flattening
• Activation Functions: ReLU / Softmax
• 10 epochs
• 90% accuracy achieved on unseen data

Project 2 : IMBD Review Text Classification (Positive/Negative)

Classify IMBD movie review as either positive or negative

• 25,000 training images, 25,000 test images
• In progress...