human_gesture_classify

GOAL

• Classify gestures by reading human hand muscle activity.

DATASET

• Provided in the url: https://www.kaggle.com/kyr7plus/emg-4
• Number of Classes: 4
• Class Labels: rock, scissors, paper, ok

REQUIREMENTS

• Explain each component in the design cycle of pattern recognition system.
• Compare results using 3 different classifiers.
• Use 5-fold cross validation or the holdout method ( %70 for training and %30 for testing) for the evaluation.
• Employ a proper feature selection method to determine (sub)optimal feature subset.
• Provide confusion matrix and F-scores.