Fall Detection System using XGBoost and IoT

Abstract

This project aims to design and implement a fall detection system for the elders using machine learning techniques and Internet-of-Things (IoT). The main issue with fall detection systems is false alarms and hence incorporating machine learning in the fall detection algorithm can tackle this problem. Therefore, choosing the right machine learning algorithm for the given problem is essential and several factors need to be considered in making that choice. For this project, the XGBoost algorithm is used and the machine learning model is trained on the Sisfall dataset. A wearable device that is worn on the waist is designed using an accelerometer, a microcontroller, a Global Positioning System (GPS) module and a buzzer. The acceleration data obtained is converted into features and fed into the machine learning model which will then make a prediction. If a fall event has occurred, the buzzer is activated and emergency contacts of the victim are notified immediately using IoT and Global System for Mobile Communications (GSM). This allows the fall victim to be attended quickly, thus reducing the negative consequences of the fall. The details of the fall are stored on the cloud so that they can be easily accessed by healthcare professionals.Testing the system concluded that the XGBoost machine learning algorithm is well suited for this problem due to the small percentage error obtained.

Clustering analysis and machine learning algorithms in the prediction of dietary patterns: Cross‐sectional results of the Brazilian Longitudinal Study of Adult Health (ELSA‐Brasil)

“…We also used the eXtreme Gradient Boosting package (XGBoost), which is an efficient and popular implementation of gradient boosted decision trees. XGBoost strictly prioritises computational speed and model performance with good accuracy using most data sets 43 . To identify the interactions among the variables, the Explain Interactions in the XGBoost (EIX) package was used.…”

“…XGBoost strictly prioritises computational speed and model performance with good accuracy using most data sets. 43 To identify the interactions among the variables, the Explain Interactions in the XGBoost (EIX) package was used. We created a ranking of the interactions using the function importance with the parameter option 'interactions'.…”

Fall Detection Algorithm Based on Inertial Sensor and Hierarchical Decision

“…In 800 falls and 1000 activities of daily living, the overall average accuracy of the fall detection system reached 99.30%, and the false positive rate was lower than 0.69%. Cahoolessur et al [ 15 ] used the XGBoost algorithm and trained a machine learning model on the Sisfall dataset. Wearable devices worn at the waist use an accelerometer, microcontroller, global positioning system (GPS) module, and a buzzer design.…”

A Simplistic and Cost-Effective Design for Real-World Development of an Ambient Assisted Living System for Fall Detection and Indoor Localization: Proof of Concept

“…This gadget had a three-part assembly: a camera, a gyroscope, and an accelerometer, all of which were connected to a computer that comprised the system architecture to detect falls. The work of Cahoolessur et al in [46] introduced a binary classifier-based device capable of finding anomalies in behavioral patterns such as falls in a simulated IoT-based environment. To design the wearable gadget, the authors first developed a model for the user, using a cloud computing-based architecture, which was followed by implementation and testing of the same.…”

“…6. Some of the works have also involved the development of new applicationssuch as the smartphone-based application proposed in [30] and the wearable devices proposed in [46], [49], and [50]. Replicating the design of an application has several challenges unless it is replicated or re-developed by the original developers [82].…”