Human Activity Recognition for Multi-label Classification in Smart Homes Using Ensemble Methods

Author(s):  
John W. Kasubi ◽  
Manjaiah D. Huchaiah
2017 ◽  
Vol 13 (2) ◽  
pp. 58-78 ◽  
Author(s):  
Samaneh Zolfaghari ◽  
Mohammad Reza Keyvanpour ◽  
Raziyeh Zall

New advancements in pervasive computing technology have turned smart homes into a daily living monitoring tool increasingly used for elderly. Recently, using knowledge driven approaches such as ontology to introduce semantic smart homes has received attention due to their flexibility, reasoning and knowledge representation. Due to the vast number of ontological human activity recognition methods, the proposed ontological human activity recognition framework can be effective in analyzing and evaluating different methods in different applications and dealing with various challenges. Also, due to numerous challenges involved in different aspects of ontology-based human activity recognition in smart homes, this paper offers a classification for challenges in human activity recognition in ontology based systems. Then the proposed ontological human activity recognition framework is evaluated based on the proposed classification and ontology-based techniques which are thought to solve some of the challenges are examined and analyzed.


2020 ◽  
Vol 140 ◽  
pp. 112849
Author(s):  
Amina Jarraya ◽  
Amel Bouzeghoub ◽  
Amel Borgi ◽  
Khedija Arour

Sensors ◽  
2019 ◽  
Vol 19 (20) ◽  
pp. 4474 ◽  
Author(s):  
Du ◽  
Lim ◽  
Tan

Smart Homes are generally considered the final solution for living problem, especially for the health care of the elderly and disabled, power saving, etc. Human activity recognition in smart homes is the key to achieving home automation, which enables the smart services to automatically run according to the human mind. Recent research has made a lot of progress in this field; however, most of them can only recognize default activities, which is probably not needed by smart homes services. In addition, low scalability makes such research infeasible to be used outside the laboratory. In this study, we unwrap this issue and propose a novel framework to not only recognize human activity but also predict it. The framework contains three stages: recognition after the activity, recognition in progress, and activity prediction in advance. Furthermore, using passive RFID tags, the hardware cost of our framework is sufficiently low to popularize the framework. In addition, the experimental result demonstrates that our framework can realize good performance in both activity recognition and prediction with high scalability.


Sensors ◽  
2019 ◽  
Vol 19 (9) ◽  
pp. 2039 ◽  
Author(s):  
Yiming Tian ◽  
Xitai Wang ◽  
Lingling Chen ◽  
Zuojun Liu

Sensor-based human activity recognition can benefit a variety of applications such as health care, fitness, smart homes, rehabilitation training, and so forth. In this paper, we propose a novel two-layer diversity-enhanced multiclassifier recognition method for single wearable accelerometer-based human activity recognition, which contains data-based and classifier-based diversity enhancement. Firstly, we introduce the kernel Fisher discriminant analysis (KFDA) technique to spatially transform the training samples and enhance the discrimination between activities. In addition, bootstrap resampling is utilized to increase the diversities of the dataset for training the base classifiers in the multiclassifier system. Secondly, a combined diversity measure for selecting the base classifiers with excellent performance and large diversity is proposed to optimize the performance of the multiclassifier system. Lastly, majority voting is utilized to combine the preferred base classifiers. Experiments showed that the data-based diversity enhancement can improve the discriminance of different activity samples and promote the generation of base classifiers with different structures and performances. Compared with random selection and traditional ensemble methods, including Bagging and Adaboost, the proposed method achieved 92.3% accuracy and 90.7% recall, which demonstrates better performance in activity recognition.


Sign in / Sign up

Export Citation Format

Share Document