Patient falls in hospitals and nursing homes are often caused by unsupervised activities such as getting out of bed and walking around in unsteady manner, leading to the need for increased supervision as a fall prevention strategy. The development of battery-free, lightweight, and wearable sensors is opening up new opportunities for ambulatory monitoring, especially for older people, as these sensors are non-intrusive and well-suited for monitoring their activities. Unobtrusive design makes them particularly suitable for monitoring seniors. The data set involves motion data collected from 14 elderly individuals, ranging in age from 66 to 86 years, using a wearable sensor that doesn't require batteries. The purpose of the data collection is to identify different activities carried out by these individuals in a clinical setting. The wearable sensor was placed on top of their clothing during the data collection. This project involves the development of a python program using machine learning. Machine learning focuses on making predictions for data that has not been seen before, referred to as testing data. A collection of algorithms is necessary to carry out this task in the field of machine learning. We applied 4 different algorithms to our dataset and found that random forest gives us the best model accuracy, i.e., 99.3%. Then we normalized and standardized the data our data to check if the accuracy of the models goes up. As expected, the accuracy of all models increased by a noticeable amount when the data is standardized. KNN’s accuracy is increased by 7%. Logistic regression’s accuracy is improved by 31% when the data is standardized rather than normalized. SVM’s accuracy increased by 39%. Random forest accuracy is unchanged even when the data is normalized or standardized. The best accuracies of all models are KNN (standardized)-98.8%, Logistic Regression (standardized)-90.9%, Random Forest (any)-99.3%, and SVM (standardized)-97.9%.