Trip detection method based on characteristic weighting and improved Bayesian algorithm

Abstract

The invention relates to a trip detection method based on characteristic weighting and an improved Bayesian algorithm. The method comprises the operation steps: 1), trip modeling; 2), characteristic selection and characteristic weighting; 3), trip detection based on the improved Bayesian algorithm. The machine learning method based on a naive Bayesian framework carries out the improvement of the trip detection, enables a system to be intelligent, and enables the system to be able to meet the requirements of trip detection of individuals with different characteristics. The method gives consideration to the importance of different characteristics in a trip detection process, combines a feature selection method and the characteristic weighting, finally improves a naive Bayesian algorithm according to the characteristic weighting, and improves the detection performance.

Description

technical field [0001] The invention relates to the technical field of health monitoring, in particular to a fall detection method based on feature weighting and improved Bayesian algorithm. Background technique [0002] Today, as the world's population ages, it is increasingly common for older people to live alone. Due to the degeneration of the physiological functions of the elderly, falls have become a prominent problem for the elderly. Most falls will cause physical injury to the elderly, and psychological pressure will be generated due to the lack of timely assistance after the fall, leading to serious consequences. Therefore, timely and accurate detection of the fall behavior of the elderly is the main goal of the fall detection system. The fall detection algorithm is the core part of the fall detection system. In the current research, most fall detection systems use a threshold-based method to detect falls, which cannot match the human body with different characteri...

AI Technical Summary

Problems solved by technology

The fall detection algorithm is the core part of the fall detection system. In the current research, most fall detection systems use a threshold-based method to detect falls, which cannot match the human body with different characteristics well. Physical states with similar processes have a higher misjudgment rate during detection

In order to improve the accuracy of fall detection, some systems extract a large number of features to detect falls, making redundant features increase the feature dimension, resulting in complex system operations

Images