Irregularity Detection of Daily Behavior Patterns Based on Unsupervised Learning

摘要

The irregularity detection of daily behaviors has received lots of attention, especially in homecare. An IRregularity Detection (IRD) algorithm is proposed to identify the irregular behavior patterns using the unsupervised learning. The distance and similarity between daily behavior patterns are designed as important features to build up the irregularity detection model. Experiments demonstrate that the proposed algorithm exceeds the existing unsupervised machine learning algorithms in terms of the numbers of False Negative and False Positive.