This study evaluated six machine learning classifiers for human activity recognition using smartphone sensor data. The classifiers - Decision Stump, Hoeffding Tree, Random Tree, J48, Random Forest, and REP Tree - were tested on a publicly available dataset containing accelerometer readings for six activities. Random Forest, J48, and REP Tree achieved statistically significantly better accuracy than Decision Stump and Hoeffding Tree according to measures like overall accuracy, precision, recall, F-measure, and kappa statistic. The results suggest these three classifiers are well-suited for human activity recognition from smartphone sensor data.