Pub Date : 2017-12-01DOI: 10.1109/ICMLA.2017.0-106
M. Ogihara, Gang Ren
This paper investigates the use of linguistic features extracted from the application essays of students enrolled in a university academic program for their retention pattern prediction. Three sets of linguistic features are generated from text analysis: (1) latent Dirichlet allocation (LDA) based topic modeling with a variety of topic numbers, (2) Linguistic Inquiry and Word Count (LIWC), and (3) part-of-speech (POS) distribution. Various classification experiments are implemented to evaluate the prediction performance of student retention patterns from these three feature sets and their combinations. The results show that the POS distribution features yield the best prediction performance among these three, while neither the LDA features nor ensemble methods improves predictive performance, which is contrary to admission experts’ manual analysis methods in the conventional admission processes.
{"title":"Student Retention Pattern Prediction Employing Linguistic Features Extracted from Admission Application Essays","authors":"M. Ogihara, Gang Ren","doi":"10.1109/ICMLA.2017.0-106","DOIUrl":"https://doi.org/10.1109/ICMLA.2017.0-106","url":null,"abstract":"This paper investigates the use of linguistic features extracted from the application essays of students enrolled in a university academic program for their retention pattern prediction. Three sets of linguistic features are generated from text analysis: (1) latent Dirichlet allocation (LDA) based topic modeling with a variety of topic numbers, (2) Linguistic Inquiry and Word Count (LIWC), and (3) part-of-speech (POS) distribution. Various classification experiments are implemented to evaluate the prediction performance of student retention patterns from these three feature sets and their combinations. The results show that the POS distribution features yield the best prediction performance among these three, while neither the LDA features nor ensemble methods improves predictive performance, which is contrary to admission experts’ manual analysis methods in the conventional admission processes.","PeriodicalId":6636,"journal":{"name":"2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"15 11","pages":"532-539"},"PeriodicalIF":0.0,"publicationDate":"2017-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91521356","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-12-01DOI: 10.1109/ICMLA.2017.00-74
Jianping Fan, Tianyi Zhao, Zhenzhong Kuang, Zhou Yu, Jun Yu
In this paper, a deep mixture algorithm is developed to support large-scale visual recognition (e.g., recognizing tens of thousands of object classes) by seamlessly combining a set of base deep CNNs (AlexNet) with diverse task spaces, e.g., such base deep CNNs (i.e., diverse experts) are trained to recognize different subsets of tens of thousands of object classes rather than the same set of object classes. Our experimental results have demonstrated that our deep mixture algorithm can achieve very competitive results on large-scale visual recognition.
{"title":"Deep Mixture of Experts with Diverse Task Spaces","authors":"Jianping Fan, Tianyi Zhao, Zhenzhong Kuang, Zhou Yu, Jun Yu","doi":"10.1109/ICMLA.2017.00-74","DOIUrl":"https://doi.org/10.1109/ICMLA.2017.00-74","url":null,"abstract":"In this paper, a deep mixture algorithm is developed to support large-scale visual recognition (e.g., recognizing tens of thousands of object classes) by seamlessly combining a set of base deep CNNs (AlexNet) with diverse task spaces, e.g., such base deep CNNs (i.e., diverse experts) are trained to recognize different subsets of tens of thousands of object classes rather than the same set of object classes. Our experimental results have demonstrated that our deep mixture algorithm can achieve very competitive results on large-scale visual recognition.","PeriodicalId":6636,"journal":{"name":"2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"1 1","pages":"721-725"},"PeriodicalIF":0.0,"publicationDate":"2017-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88669969","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-12-01DOI: 10.1109/ICMLA.2017.0-131
O. Dehzangi, Muhamed Farooq
The use of Brain Computer Interface (BCI) systems in the Intensive Care Unit (ICU) can facilitate communication on demand. BCI systems enable ICU patients to communicate using the electrical activity of their brains. For this purpose, we designed and developed a BCI system comprised of an Android tablet that allows patients to look at the screen to ask for what they need using their Electroencephalogram (EEG) recorded using a wireless wearable BCI. However, there are two main challenges associated with the BCI application. Due to the insufficient screen refresh rate of the mobile device, the flickering stimuli is imprecise. Hence, we introduce a partition-based feature extraction and fusion method using Canonical Correlation Analysis (CCA) and Power Spectral Density Analysis (PSDA) to overcome this limitation. Also, BCI devices require a calibration stage in order to capture subject-specific information, which might be particularly troublesome for ICU patients. WE hypothesize that inducing subject related information in the model training and adaptation improves the overall SSVEP identification performance with minimal calibration requirements. As such, We propose a three stage Gaussian Mixture Model (GMM)-based model training and subject adaptation: 1) we generate a subject independent universal GMM model, 2) we generate subject-dependent identification models using only a few collected SSVEP segments from each patient, and 3) we form a vector out of the subject-dependent GMMs and pass it to Support Vector Machine (SVM) for SSVEP target frequency identification. Our experimental results on 10 subjects demonstrated that the proposed framework yielded very efficient SSVEP identification performances achieving 98.7% accuracy using our most accurate model.
{"title":"Subject-Dependent SSVEP Identification Using GMM Training and Adaptation","authors":"O. Dehzangi, Muhamed Farooq","doi":"10.1109/ICMLA.2017.0-131","DOIUrl":"https://doi.org/10.1109/ICMLA.2017.0-131","url":null,"abstract":"The use of Brain Computer Interface (BCI) systems in the Intensive Care Unit (ICU) can facilitate communication on demand. BCI systems enable ICU patients to communicate using the electrical activity of their brains. For this purpose, we designed and developed a BCI system comprised of an Android tablet that allows patients to look at the screen to ask for what they need using their Electroencephalogram (EEG) recorded using a wireless wearable BCI. However, there are two main challenges associated with the BCI application. Due to the insufficient screen refresh rate of the mobile device, the flickering stimuli is imprecise. Hence, we introduce a partition-based feature extraction and fusion method using Canonical Correlation Analysis (CCA) and Power Spectral Density Analysis (PSDA) to overcome this limitation. Also, BCI devices require a calibration stage in order to capture subject-specific information, which might be particularly troublesome for ICU patients. WE hypothesize that inducing subject related information in the model training and adaptation improves the overall SSVEP identification performance with minimal calibration requirements. As such, We propose a three stage Gaussian Mixture Model (GMM)-based model training and subject adaptation: 1) we generate a subject independent universal GMM model, 2) we generate subject-dependent identification models using only a few collected SSVEP segments from each patient, and 3) we form a vector out of the subject-dependent GMMs and pass it to Support Vector Machine (SVM) for SSVEP target frequency identification. Our experimental results on 10 subjects demonstrated that the proposed framework yielded very efficient SSVEP identification performances achieving 98.7% accuracy using our most accurate model.","PeriodicalId":6636,"journal":{"name":"2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"86 1","pages":"384-389"},"PeriodicalIF":0.0,"publicationDate":"2017-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84795154","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-12-01DOI: 10.1109/ICMLA.2017.00-23
Caleb Vununu, Ki-Ryong Kwon, Eung-Joo Lee, Kwang-Seok Moon, Suk-Hwan Lee
Machine fault diagnosis (MFD) recovers all the studies that aim to detect faults automatically in the machines. This study aims to develop a sound based MFD system for drills using the pattern recognition techniques such as principal components analysis (PCA) and artificial neural networks (ANN). The sound signals emitted by healthy and faulty drills are obtained and analyzed. Unlike the conventional methods that focus on the time domain, we explore here the effectiveness of the frequency domain components and demonstrate the ineffectiveness of the time domain based analysis of the sounds produced by the drills. The power spectrum components of the sounds are extracted before using PCA for the purpose of dimensionality reduction and redundancy removal. The first principal components are then selected and given to a neural network based classifier in order to perform the diagnosis. The results show that the proposed method can be used for the sounds based automatic diagnosis system.
{"title":"Automatic Fault Diagnosis of Drills Using Artificial Neural Networks","authors":"Caleb Vununu, Ki-Ryong Kwon, Eung-Joo Lee, Kwang-Seok Moon, Suk-Hwan Lee","doi":"10.1109/ICMLA.2017.00-23","DOIUrl":"https://doi.org/10.1109/ICMLA.2017.00-23","url":null,"abstract":"Machine fault diagnosis (MFD) recovers all the studies that aim to detect faults automatically in the machines. This study aims to develop a sound based MFD system for drills using the pattern recognition techniques such as principal components analysis (PCA) and artificial neural networks (ANN). The sound signals emitted by healthy and faulty drills are obtained and analyzed. Unlike the conventional methods that focus on the time domain, we explore here the effectiveness of the frequency domain components and demonstrate the ineffectiveness of the time domain based analysis of the sounds produced by the drills. The power spectrum components of the sounds are extracted before using PCA for the purpose of dimensionality reduction and redundancy removal. The first principal components are then selected and given to a neural network based classifier in order to perform the diagnosis. The results show that the proposed method can be used for the sounds based automatic diagnosis system.","PeriodicalId":6636,"journal":{"name":"2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"57 1","pages":"992-995"},"PeriodicalIF":0.0,"publicationDate":"2017-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90771074","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-12-01DOI: 10.1109/ICMLA.2017.00-73
Jun Yu, Zhenzhong Kuang, Zhou Yu, D. Lin, Jianping Fan
This paper aims to simultaneously consider two inseparable issues for privacy setting recommendation: (1) sensitiveness of visual content of the images being shared; and (2) trustworthiness of users being granted. First, an object-based approach is developed for image content sensitiveness (privacy) representation. Secondly, the users on a social network are clustered into a set of representative social groups to generate a discriminative dictionary for user trustworthiness characterization. Finally, a tree classifier is trained hierarchically to recommend appropriate privacy settings for image sharing.
{"title":"Privacy Setting Recommendation for Image Sharing","authors":"Jun Yu, Zhenzhong Kuang, Zhou Yu, D. Lin, Jianping Fan","doi":"10.1109/ICMLA.2017.00-73","DOIUrl":"https://doi.org/10.1109/ICMLA.2017.00-73","url":null,"abstract":"This paper aims to simultaneously consider two inseparable issues for privacy setting recommendation: (1) sensitiveness of visual content of the images being shared; and (2) trustworthiness of users being granted. First, an object-based approach is developed for image content sensitiveness (privacy) representation. Secondly, the users on a social network are clustered into a set of representative social groups to generate a discriminative dictionary for user trustworthiness characterization. Finally, a tree classifier is trained hierarchically to recommend appropriate privacy settings for image sharing.","PeriodicalId":6636,"journal":{"name":"2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"77 1","pages":"726-730"},"PeriodicalIF":0.0,"publicationDate":"2017-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"91189180","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-12-01DOI: 10.1109/ICMLA.2017.00017
Naomi Joseph, Parita Sanghani, Hongliang Ren
Glioblastomas (GBMs) are cancerous brain tumors that require careful and intricate analysis for surgical planning. Physicians employ Magnetic Resonance Imaging (MRI) in order to diagnose glioblastomas. The segmentation of the tumor is a crucial step in surgical planning. Clinicians manually segment the tumor voxel-by-voxel; however, this is very time consuming. Hence, extensive research has been conducted to semi-automate and fully-automate this segmentation process. This project explores manual segmentation and utilizes k-means clustering technique for semi-automated segmentation. The accuracy of the k-means clustering segmentation was measured using the Dice Coefficient (DC). The results show that k-means clustering provides high accuracy for the segmentation of the enhanced region of tumor (which appears bright in the T1 post contrast MR image) and hence, it can be efficiently used to speed up manual segmentation.
{"title":"Semi-Automated Segmentation of Glioblastomas in Brain MRI Using Machine Learning Techniques","authors":"Naomi Joseph, Parita Sanghani, Hongliang Ren","doi":"10.1109/ICMLA.2017.00017","DOIUrl":"https://doi.org/10.1109/ICMLA.2017.00017","url":null,"abstract":"Glioblastomas (GBMs) are cancerous brain tumors that require careful and intricate analysis for surgical planning. Physicians employ Magnetic Resonance Imaging (MRI) in order to diagnose glioblastomas. The segmentation of the tumor is a crucial step in surgical planning. Clinicians manually segment the tumor voxel-by-voxel; however, this is very time consuming. Hence, extensive research has been conducted to semi-automate and fully-automate this segmentation process. This project explores manual segmentation and utilizes k-means clustering technique for semi-automated segmentation. The accuracy of the k-means clustering segmentation was measured using the Dice Coefficient (DC). The results show that k-means clustering provides high accuracy for the segmentation of the enhanced region of tumor (which appears bright in the T1 post contrast MR image) and hence, it can be efficiently used to speed up manual segmentation.","PeriodicalId":6636,"journal":{"name":"2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"231 1","pages":"1149-1152"},"PeriodicalIF":0.0,"publicationDate":"2017-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"78099879","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-12-01DOI: 10.1109/ICMLA.2017.00-95
Evgeni Bikov, P. Boyko, Evgeny Sokolov, D. Yarotsky
Modern railway networks include thousands of failure registration devices, and prompt response to detected failures is critical to normal network operation. However, a large share of produced alerts may be formed by false alarms associated with maintenance or faulty diagnostics, thus hindering the processing of actual failures. It is therefore very desirable to perform fast automated intelligent ranking of incidents before they are analyzed by human operators. In this paper we describe a machine-learning-based incident ranking model that we have developed and deployed at the Moscow Railway network (a large network with 500+ stations). The model estimates the probability of failure using multiple features of the incident at hand. The model was constructed using the XGBoost library and a database of 5 million historical incidents. The model shows high accuracy (AUC 0.901) in the deployment environment.
{"title":"Railway Incident Ranking with Machine Learning","authors":"Evgeni Bikov, P. Boyko, Evgeny Sokolov, D. Yarotsky","doi":"10.1109/ICMLA.2017.00-95","DOIUrl":"https://doi.org/10.1109/ICMLA.2017.00-95","url":null,"abstract":"Modern railway networks include thousands of failure registration devices, and prompt response to detected failures is critical to normal network operation. However, a large share of produced alerts may be formed by false alarms associated with maintenance or faulty diagnostics, thus hindering the processing of actual failures. It is therefore very desirable to perform fast automated intelligent ranking of incidents before they are analyzed by human operators. In this paper we describe a machine-learning-based incident ranking model that we have developed and deployed at the Moscow Railway network (a large network with 500+ stations). The model estimates the probability of failure using multiple features of the incident at hand. The model was constructed using the XGBoost library and a database of 5 million historical incidents. The model shows high accuracy (AUC 0.901) in the deployment environment.","PeriodicalId":6636,"journal":{"name":"2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"195 1","pages":"601-606"},"PeriodicalIF":0.0,"publicationDate":"2017-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74947107","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-12-01DOI: 10.1109/ICMLA.2017.0-142
Koosha Sadeghi, Ayan Banerjee, Javad Sohankar, S. Gupta
Feature extraction and Machine Learning (ML) techniques are required to reduce high variability of biometric data in Biometric Authentication Systems (BAS) toward improving system utilization (acceptance of legitimate subjects). However, reduction in data variability, also decreases the adversary’s effort in manufacturing legitimate biometric data to break the system (security strength). Typically for BAS design, security strength is evaluated through variability analysis on data, regardless of feature extraction and ML, which are essential for accurate evaluation. In this research, we provide a geometrical method to measure the security strength in BAS, which analyzes the effects of feature extraction and ML on the biometric data. Using the proposed method, we evaluate the security strength of five state-of-the-art electroencephalogram-based authentication systems, on data from 106 subjects, and the maximum achievable security strength is 83 bits.
{"title":"Geometrical Analysis of Machine Learning Security in Biometric Authentication Systems","authors":"Koosha Sadeghi, Ayan Banerjee, Javad Sohankar, S. Gupta","doi":"10.1109/ICMLA.2017.0-142","DOIUrl":"https://doi.org/10.1109/ICMLA.2017.0-142","url":null,"abstract":"Feature extraction and Machine Learning (ML) techniques are required to reduce high variability of biometric data in Biometric Authentication Systems (BAS) toward improving system utilization (acceptance of legitimate subjects). However, reduction in data variability, also decreases the adversary’s effort in manufacturing legitimate biometric data to break the system (security strength). Typically for BAS design, security strength is evaluated through variability analysis on data, regardless of feature extraction and ML, which are essential for accurate evaluation. In this research, we provide a geometrical method to measure the security strength in BAS, which analyzes the effects of feature extraction and ML on the biometric data. Using the proposed method, we evaluate the security strength of five state-of-the-art electroencephalogram-based authentication systems, on data from 106 subjects, and the maximum achievable security strength is 83 bits.","PeriodicalId":6636,"journal":{"name":"2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"46 1","pages":"309-314"},"PeriodicalIF":0.0,"publicationDate":"2017-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74306400","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-12-01DOI: 10.1109/ICMLA.2017.000-2
N. N. Qomariyah, D. Kazakov
Building a proactive and unobtrusive recom- mender system is still a challenging task. In the real world, buyers may be offered a lot of choices while trying to choose the item that best suits their preference. Such items may have many attributes, which can complicate the process. The classic approach in decision support systems – to put weights on the importance of each attribute – is not always helpful here. For instance, there are cases when users find it is hard to formulate their priorities explicitly. In this paper, we promote the use of pairwise comparisons, which allow the user preferences to be inferred rather than spell out. Our system aims to learn from a limited number of examples and using clustering to guide the selection of pairs for annotation. The approach is demonstrated in the case of purchasing a used car using a large, real-world data set.
{"title":"Mixed Type Multi-attribute Pairwise Comparisons Learning","authors":"N. N. Qomariyah, D. Kazakov","doi":"10.1109/ICMLA.2017.000-2","DOIUrl":"https://doi.org/10.1109/ICMLA.2017.000-2","url":null,"abstract":"Building a proactive and unobtrusive recom- mender system is still a challenging task. In the real world, buyers may be offered a lot of choices while trying to choose the item that best suits their preference. Such items may have many attributes, which can complicate the process. The classic approach in decision support systems – to put weights on the importance of each attribute – is not always helpful here. For instance, there are cases when users find it is hard to formulate their priorities explicitly. In this paper, we promote the use of pairwise comparisons, which allow the user preferences to be inferred rather than spell out. Our system aims to learn from a limited number of examples and using clustering to guide the selection of pairs for annotation. The approach is demonstrated in the case of purchasing a used car using a large, real-world data set.","PeriodicalId":6636,"journal":{"name":"2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"418 1","pages":"1094-1097"},"PeriodicalIF":0.0,"publicationDate":"2017-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84909195","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Pub Date : 2017-12-01DOI: 10.1109/ICMLA.2017.0-148
Han Zou, Yuxun Zhou, Jianfei Yang, Weixi Gu, Lihua Xie, C. Spanos
Human activity recognition is becoming the vital underpinning for a myriad of emerging applications in the field of human-computer interaction, mobile computing, and smart grid. Besides the utilization of up-to-date sensing techniques, modern activity recognition systems also require a machine learning (ML) algorithm that leverages the sensory data for identification purposes. In view of the unique characteristics of the measurement data and the ML challenges thereof, we propose a non-intrusive human activity recognition system that only uses existing commodity WiFi routers. The core of our system is a novel multiple kernel representation learning (MKRL) framework that automatically extracts and combines informative patterns from the Channel State Information (CSI) measurements. The MKRL firstly learns a kernel string representation from time, frequency, wavelet, and shape domains with an efficient greedy algorithm. Then it performs information fusion from diverse perspectives based on multi-view kernel learning. Moreover, different stages of MKRL can be seamlessly integrated into a multiple kernel learning framework to build up a robust and comprehensive activity classifier. Extensive experiments are conducted in typical indoor environments and the experimental results demonstrate that the proposed system outperforms existing methods and achieves a 98% activity recognition accuracy.
{"title":"Multiple Kernel Representation Learning for WiFi-Based Human Activity Recognition","authors":"Han Zou, Yuxun Zhou, Jianfei Yang, Weixi Gu, Lihua Xie, C. Spanos","doi":"10.1109/ICMLA.2017.0-148","DOIUrl":"https://doi.org/10.1109/ICMLA.2017.0-148","url":null,"abstract":"Human activity recognition is becoming the vital underpinning for a myriad of emerging applications in the field of human-computer interaction, mobile computing, and smart grid. Besides the utilization of up-to-date sensing techniques, modern activity recognition systems also require a machine learning (ML) algorithm that leverages the sensory data for identification purposes. In view of the unique characteristics of the measurement data and the ML challenges thereof, we propose a non-intrusive human activity recognition system that only uses existing commodity WiFi routers. The core of our system is a novel multiple kernel representation learning (MKRL) framework that automatically extracts and combines informative patterns from the Channel State Information (CSI) measurements. The MKRL firstly learns a kernel string representation from time, frequency, wavelet, and shape domains with an efficient greedy algorithm. Then it performs information fusion from diverse perspectives based on multi-view kernel learning. Moreover, different stages of MKRL can be seamlessly integrated into a multiple kernel learning framework to build up a robust and comprehensive activity classifier. Extensive experiments are conducted in typical indoor environments and the experimental results demonstrate that the proposed system outperforms existing methods and achieves a 98% activity recognition accuracy.","PeriodicalId":6636,"journal":{"name":"2017 16th IEEE International Conference on Machine Learning and Applications (ICMLA)","volume":"422 1","pages":"268-274"},"PeriodicalIF":0.0,"publicationDate":"2017-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84932830","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：