Activity recognition has become the necessity of smart homes, future factories, and surveillance. Activities independent of body posture predominantly exhibiting gestures involving both arm and the wrist motion supports the use of the wearable sensors for data acquisition. This paper uses an algorithm based prediction method to recognize the Activities of Daily Life (ADL) involving activities like mobility, feeding, and functional transfers. The classification of the various activities were carried out by using decision tree - J48 algorithm from the acquired dataset.