Two phase ensemble classifier for smartphone based human activity recognition independent of hardware configuration and usage behaviour

Saha, Jayita; Chowdhury, Chandreyee; Biswas, Suparna

doi:10.1007/s00542-018-3802-9

Cited by 23 publications

(14 citation statements)

References 43 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…5. Extensive experiments to explore the effectiveness of the proposed methods using two publicly available datasets and compare the significance of the multi-view stacking ensemble with weighted majority voting, Bagging and Random Subspace ensemble [16,26,27] based multiple classifier system methods.…”

Section: To Demonstrate the Impact Of Synthetic Minority Over-samplinmentioning

confidence: 99%

“…The data were processed and learned separately and combined at the classifier level in order to achieve generalizability and independence of different activity contexts. While Saha et al [16] propose a two-phase ensemble algorithm for human activity recognition by exploiting position specific condition to improve performance results. Therefore, the training and testing data were drawn from different placement positions.…”

Section: Multiple Classifier Systemsmentioning

confidence: 99%

“…To provide enhanced performance and diversity in recognition of human activity details, multi-sensor fusion strategies have been proposed in recent studies [15,16], in which various sensor modalities outlined earlier are integrated using raw sensor signal, extracted features or decisions predicted by individual classification algorithms. The central ideas of fusion protocols are to incorporate diverse sensor modalities to increase reliabilities, robustness and enhanced generalization of human activity detection frameworks.…”

Section: Introductionmentioning

confidence: 99%

“…However, few studies [16,19,20] have address the issue of multiple classifier system methods for human activity identification by utilizing diverse multimodal sensor data and classification algorithms. Specifically, these studies only developed protocols to integrate multiple accelerometer sensors attached at different body locations, which limit their implementation for robust activity recognition.…”

Section: Introductionmentioning

confidence: 99%

See 3 more Smart Citations

Multi-sensor fusion based on multiple classifier systems for human activity identification

Nweke

Wah

Mujtaba

et al. 2019

Hum. Cent. Comput. Inf. Sci.

View full text Add to dashboard Cite

Multimodal sensors in healthcare applications have been increasingly researched because it facilitates automatic and comprehensive monitoring of human behaviors, high-intensity sports management, energy expenditure estimation, and postural detection. Recent studies have shown the importance of multi-sensor fusion to achieve robustness, high-performance generalization, provide diversity and tackle challenging issue that maybe difficult with single sensor values. The aim of this study is to propose an innovative multi-sensor fusion framework to improve human activity detection performances and reduce misrecognition rate. The study proposes a multiview ensemble algorithm to integrate predicted values of different motion sensors. To this end, computationally efficient classification algorithms such as decision tree, logistic regression and k-Nearest Neighbors were used to implement diverse, flexible and dynamic human activity detection systems. To provide compact feature vector representation, we studied hybrid bio-inspired evolutionary search algorithm and correlation-based feature selection method and evaluate their impact on extracted feature vectors from individual sensor modality. Furthermore, we utilized Synthetic Over-sampling minority Techniques (SMOTE) algorithm to reduce the impact of class imbalance and improve performance results. With the above methods, this paper provides unified framework to resolve major challenges in human activity identification. The performance results obtained using two publicly available datasets showed significant improvement over baseline methods in the detection of specific activity details and reduced error rate. The performance results of our evaluation showed 3% to 24% improvement in accuracy, recall, precision, F-measure and detection ability (AUC) compared to single sensors and feature-level fusion. The benefit of the proposed multisensor fusion is the ability to utilize distinct feature characteristics of individual sensor and multiple classifier systems to improve recognition accuracy. In addition, the study suggests a promising potential of hybrid feature selection approach, diversity-based multiple classifier systems to improve mobile and wearable sensor-based human activity detection and health monitoring system.

show abstract

Section: To Demonstrate the Impact Of Synthetic Minority Over-samplinmentioning

confidence: 99%

Section: Multiple Classifier Systemsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Multi-sensor fusion based on multiple classifier systems for human activity identification

Nweke

Wah

Mujtaba

et al. 2019

Hum. Cent. Comput. Inf. Sci.

View full text Add to dashboard Cite

show abstract

“…In terms of behavior detection, there are many smartphone-based researches using the high frequency transient data captured by embedded sensors. By placing the smartphone in a relatively fixed position (such as in a pocket) of an operator, some features such as the size and the frequency of the wave peaks and the roughness, can be recorded with the tri-axial accelerometer [13][14][15][16] .…”

Section: Introduction mentioning

confidence: 99%

Behavior modelling and sensing for machinery operations using smartphone’s sensor data: A case study of forage maize sowing

Wu¹,

Chen²,

Wang³

et al. 2019

International Journal of Agricultural and Biological Engineering

View full text Add to dashboard Cite

Large-scale agricultural machinery cooperatives require technical statistic report of agricultural machinery operations to improve the efficiency of fleet management. This research proposed a smartphone-based solution to build the behavior model for agricultural machinery operations by using the embedded sensors including the GNSS, the accelerometer, and the microphone. The whole working process of agricultural machinery operation was divided into four stages: preparation, operation, U-turn, and transfer, each of which may contain the behaviors of stalling and idling. Field experiments were carried out by skilled operators, whose operations were typical agricultural machinery operations that could be used to extract behavior features. Butterworth low-pass filter was used to smooth the output from the accelerometer. Then, the operating data were collected through an APP when sowing the forage maize as a case study. Four stages of machinery operation can be preliminarily classified by using GNSS speed, while the identification of behaviors such as sudden acceleration and longtime idling that may increase fuel consumption, reduce machinery life, or decrease the working efficiency, requires extra information such as acceleration and sound intensity. The results showed that the jerk of accelerating can describe the severity of the sudden acceleration, the standard deviation of forward acceleration can reflect the smoothness of operation, the upward acceleration can be used to identify behaviors of stalling and idling, and the sound intensity during idling can capture the behavior of goosing the throttle. Further, the operating behavior figure can be drawn based on the above parameters. In conclusion, this research constructed several behavior models of agricultural machinery and operators by using smartphone's sensor data and established the base of the online assessing and scoring system for agricultural machinery operations. modelling and sensing for machinery operations using smartphone's sensor data: a case study of forage maize sowing. Int J Agric & Biol Eng, 2019; 12(6): 66-74. Biographies: Caicong Wu, PhD, Associate Professor, research interest: location-based service of farm machinery, Email: wucc@cau.edu.cn; Zhibo Chen, PhD candidate, research interest: location-based service of farm machinery, Email: bobodi@qq.com; Dongxu Wang, Master candidate, research interest: location-based service of farm machinery, Email: dongxu_wang@ cau.edu.cn; Zhihong Kou, Master candidate, research interest: location-based service of farm machinery, Email: 603040848@qq.com; Yaping Cai, PhD candidate, research interest: mobile information collection, Email: caicai_pku@ 163.com.

show abstract

Review of Machine Learning and Deep Learning Based Recommender Systems for Health Informatics

Saha

Chowdhury

Biswas

2019

Studies in Big Data

View full text Add to dashboard Cite

Two phase ensemble classifier for smartphone based human activity recognition independent of hardware configuration and usage behaviour

Cited by 23 publications

References 43 publications

Multi-sensor fusion based on multiple classifier systems for human activity identification

Multi-sensor fusion based on multiple classifier systems for human activity identification

Behavior modelling and sensing for machinery operations using smartphone’s sensor data: A case study of forage maize sowing

Review of Machine Learning and Deep Learning Based Recommender Systems for Health Informatics

Contact Info

Product

Resources

About