AI course project
• Use Naive Bayes, MIRA and Perceptron classifiers
• Designed and Implemented 4 features including edge detection, contiguous regions, top-bottom pixel density and symmetry for digits and faces dataset.
• Accomplished 82% prediction accuracy for digits and 92% accuracy for faces detection