In this paper, we present an approach for activity recognition by using 3D skeleton data obtained by Kinect sensor. Primarily, we use the simplified Dynamic Time Wrapping(DTW) and calculate Euclidean geometry distance to obtain the probable activities from the trained data. Afterwards, for each activity, we define a modified activity feature descriptor using the interrelation of correlated joints in each frame. Before classification, we employ normalization to avoid nonuniformity in coordinates, and Principal Component Analysis(PCA) to deduce redundancy and decrease the dimensionality. Finally we apply Na?ve-Bayes-Nearest-Neighbor(NBNN) classifier to classify multiple actions. The experimental result on benchmark dataset shows that the accuracy approximates the state-of-the-art methods.