A method for unsupervised disaggregation of appliance signatures from smart meter data is presented. The primary feature used for unsupervised learning relates to abrupt transitions or magnitude changes in the power waveform. The method consists of a sequence of procedures for appliance signature identification, and disaggregation using hidden Markov modeling (HMM), and residual analysis. The key contributions are (a) a novel 'segmented' application of the Viterbi algorithm for sequence decoding with the HMM, (b) details of establishing observation and state transition probabilities for the HMM, and (c) procedures for careful handling of low power signatures. Results show that the method is effective for magnitude-based disaggregation, and provide insights for a more complete solution.
{"title":"Unsupervised Disaggregation for Non-intrusive Load Monitoring","authors":"S. Pattem","doi":"10.1109/ICMLA.2012.249","DOIUrl":"https://doi.org/10.1109/ICMLA.2012.249","url":null,"abstract":"A method for unsupervised disaggregation of appliance signatures from smart meter data is presented. The primary feature used for unsupervised learning relates to abrupt transitions or magnitude changes in the power waveform. The method consists of a sequence of procedures for appliance signature identification, and disaggregation using hidden Markov modeling (HMM), and residual analysis. The key contributions are (a) a novel 'segmented' application of the Viterbi algorithm for sequence decoding with the HMM, (b) details of establishing observation and state transition probabilities for the HMM, and (c) procedures for careful handling of low power signatures. Results show that the method is effective for magnitude-based disaggregation, and provide insights for a more complete solution.","PeriodicalId":157399,"journal":{"name":"2012 11th International Conference on Machine Learning and Applications","volume":"68 5","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"120844123","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}
Multi-label learning in graph-based relational data has gained popularity in recent years due to the increasingly complex structures of real world applications. Collective Classification deals with the simultaneous classification of neighboring instances in relational data, until a convergence criterion is reached. The rationale behind collective classification stems from the fact that an entity in a network (or relational data) is most likely influenced by the neighboring entities, and can be classified accordingly, based on the class assignment of the neighbors. Although extensive work has been done on collective classification of single labeled data, the domain of multi-labeled relational data has not been sufficiently explored. In this paper, we propose a neighborhood ranking method for multi-label classification, which can be further used in the Multi-label Collective Classification framework. We test our methods on real world datasets and also discuss the relevance of our approach for other multi-labeled relational data. Our experimental results show that the use of ranking in neighborhood selection for collective classification improves the performance of the classifier.
{"title":"Multi-label Collective Classification Using Adaptive Neighborhoods","authors":"Tanwistha Saha, H. Rangwala, C. Domeniconi","doi":"10.1109/ICMLA.2012.77","DOIUrl":"https://doi.org/10.1109/ICMLA.2012.77","url":null,"abstract":"Multi-label learning in graph-based relational data has gained popularity in recent years due to the increasingly complex structures of real world applications. Collective Classification deals with the simultaneous classification of neighboring instances in relational data, until a convergence criterion is reached. The rationale behind collective classification stems from the fact that an entity in a network (or relational data) is most likely influenced by the neighboring entities, and can be classified accordingly, based on the class assignment of the neighbors. Although extensive work has been done on collective classification of single labeled data, the domain of multi-labeled relational data has not been sufficiently explored. In this paper, we propose a neighborhood ranking method for multi-label classification, which can be further used in the Multi-label Collective Classification framework. We test our methods on real world datasets and also discuss the relevance of our approach for other multi-labeled relational data. Our experimental results show that the use of ranking in neighborhood selection for collective classification improves the performance of the classifier.","PeriodicalId":157399,"journal":{"name":"2012 11th International Conference on Machine Learning and Applications","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126576717","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}
This study presents a novel adaptive control based on a neural network for dc - dc converters. The control method is required to adapt to changes of conditions to obtain high performance dc-dc converters. In this study, the neural network control is adopted to improve the transient response of dc-dc converters. It woks in coordination with a conventional PID control to realize a high adaptive method. The neural network is trained with data which is obtained on-line. Therefore, the neural network control can adapt dynamically to change of input. The adaptation is realized by the modification of the reference in the PID control. The effect of the presented method is confirmed in simulations. Results show the presented method contributes to realize such adaptive control.
{"title":"A Novel Neural Network Based Control Method with Adaptive On-Line Training for DC-DC Converters","authors":"H. Maruta, M. Motomura, F. Kurokawa","doi":"10.1109/ICMLA.2012.152","DOIUrl":"https://doi.org/10.1109/ICMLA.2012.152","url":null,"abstract":"This study presents a novel adaptive control based on a neural network for dc - dc converters. The control method is required to adapt to changes of conditions to obtain high performance dc-dc converters. In this study, the neural network control is adopted to improve the transient response of dc-dc converters. It woks in coordination with a conventional PID control to realize a high adaptive method. The neural network is trained with data which is obtained on-line. Therefore, the neural network control can adapt dynamically to change of input. The adaptation is realized by the modification of the reference in the PID control. The effect of the presented method is confirmed in simulations. Results show the presented method contributes to realize such adaptive control.","PeriodicalId":157399,"journal":{"name":"2012 11th International Conference on Machine Learning and Applications","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126710345","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}
R. Passonneau, Ashish Tomar, Somnath Sarkar, Haimonti Dutta, Axinia Radeva
We assess the impact of an inspection repair program administered to the secondary electrical grid in New York City. The question of interest is whether repairs reduce the incidence of future events that cause service disruptions ranging from minor to serious ones. A key challenge in defining treatment and control groups in the absence of a randomized experiment involved an inherent bias in selection of electrical structures to be inspected in a given year. To compensate for the bias, we construct separate models for each year of the propensity for a structure to have an inspection repair. The propensity models account for differences across years in the structures that get inspected. To model the treatment outcome, we use a statistical approach based on the additive effects of many weak learners. Our results indicate that inspection repairs are more beneficial earlier in the five-year inspection cycle, which accords with the inherent bias to inspect structures in earlier years that are known to have problems.
{"title":"Multivariate Assessment of a Repair Program for a New York City Electrical Grid","authors":"R. Passonneau, Ashish Tomar, Somnath Sarkar, Haimonti Dutta, Axinia Radeva","doi":"10.1109/ICMLA.2012.208","DOIUrl":"https://doi.org/10.1109/ICMLA.2012.208","url":null,"abstract":"We assess the impact of an inspection repair program administered to the secondary electrical grid in New York City. The question of interest is whether repairs reduce the incidence of future events that cause service disruptions ranging from minor to serious ones. A key challenge in defining treatment and control groups in the absence of a randomized experiment involved an inherent bias in selection of electrical structures to be inspected in a given year. To compensate for the bias, we construct separate models for each year of the propensity for a structure to have an inspection repair. The propensity models account for differences across years in the structures that get inspected. To model the treatment outcome, we use a statistical approach based on the additive effects of many weak learners. Our results indicate that inspection repairs are more beneficial earlier in the five-year inspection cycle, which accords with the inherent bias to inspect structures in earlier years that are known to have problems.","PeriodicalId":157399,"journal":{"name":"2012 11th International Conference on Machine Learning and Applications","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125457799","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}
This work proposes a fast background learning algorithm for foreground detection under changing illumination. Gaussian Mixture Model (GMM) is an effective statistical model in background learning. We first focus on Titterington's online EM algorithm that can be used for real-time unsupervised GMM learning, and then advocate a deterministic data assignment strategy to avoid Bayesian computation. The color of the foreground is apt to be influenced by the environmental illumination that usually produce undesirable effect for GMM updating, however, a collinear feature of pixel intensity under changing light is discovered in RGB color space. This feature is afterward used as a reliable clue to decide which part of mixture to update under changing light. A foreground detection step proposed in early version of this work is employed to extract foreground objects by comparing the estimated background model with the current video frame. Experiments have shown the proposed method is able to achieve satisfactory static background images of scenes as well as is also superior to some mainstream methods in detection performance under both indoor and outdoor scenes.
{"title":"Real-Time Statistical Background Learning for Foreground Detection under Unstable Illuminations","authors":"Dawei Li, Lihong Xu, E. Goodman","doi":"10.1109/ICMLA.2012.85","DOIUrl":"https://doi.org/10.1109/ICMLA.2012.85","url":null,"abstract":"This work proposes a fast background learning algorithm for foreground detection under changing illumination. Gaussian Mixture Model (GMM) is an effective statistical model in background learning. We first focus on Titterington's online EM algorithm that can be used for real-time unsupervised GMM learning, and then advocate a deterministic data assignment strategy to avoid Bayesian computation. The color of the foreground is apt to be influenced by the environmental illumination that usually produce undesirable effect for GMM updating, however, a collinear feature of pixel intensity under changing light is discovered in RGB color space. This feature is afterward used as a reliable clue to decide which part of mixture to update under changing light. A foreground detection step proposed in early version of this work is employed to extract foreground objects by comparing the estimated background model with the current video frame. Experiments have shown the proposed method is able to achieve satisfactory static background images of scenes as well as is also superior to some mainstream methods in detection performance under both indoor and outdoor scenes.","PeriodicalId":157399,"journal":{"name":"2012 11th International Conference on Machine Learning and Applications","volume":"12 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126715755","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}
T. Khoshgoftaar, D. Dittman, Randall Wald, Alireza Fazelpour
Dimensionality reduction techniques have become a required step when working with bioinformatics datasets. Techniques such as feature selection have been known to not only improve computation time, but to improve the results of experiments by removing the redundant and irrelevant features or genes from consideration in subsequent analysis. Univariate feature selection techniques in particular are well suited for the large levels of high dimensionality that are inherent in bioinformatics datasets (for example: DNA microarray datasets) due to their intuitive output (a ranked lists of features or genes) and their relatively small computational time compared to other techniques. This paper presents seven univariate feature selection techniques and collects them into a single family entitled First Order Statistics (FOS) based feature selection. These seven all share the trait of using first order statistical measures such as mean and standard deviation, although this is the first work to relate them to one another and consider their performance compared with one another. In order to examine the properties of these seven techniques we performed a series of similarity and classification experiments on eleven DNA microarray datasets. Our results show that in general, each feature selection technique will create diverse feature subsets when compared to the other members of the family. However when we look at classification we find that, with one exception, the techniques will produce good classification results and that the techniques will have similar performances to each other. Our recommendation, is to use the rankers Signal-to-Noise and SAM for the best classification results and to avoid Fold Change Ratio as it is consistently the worst performer of the seven rankers.
{"title":"First Order Statistics Based Feature Selection: A Diverse and Powerful Family of Feature Seleciton Techniques","authors":"T. Khoshgoftaar, D. Dittman, Randall Wald, Alireza Fazelpour","doi":"10.1109/ICMLA.2012.192","DOIUrl":"https://doi.org/10.1109/ICMLA.2012.192","url":null,"abstract":"Dimensionality reduction techniques have become a required step when working with bioinformatics datasets. Techniques such as feature selection have been known to not only improve computation time, but to improve the results of experiments by removing the redundant and irrelevant features or genes from consideration in subsequent analysis. Univariate feature selection techniques in particular are well suited for the large levels of high dimensionality that are inherent in bioinformatics datasets (for example: DNA microarray datasets) due to their intuitive output (a ranked lists of features or genes) and their relatively small computational time compared to other techniques. This paper presents seven univariate feature selection techniques and collects them into a single family entitled First Order Statistics (FOS) based feature selection. These seven all share the trait of using first order statistical measures such as mean and standard deviation, although this is the first work to relate them to one another and consider their performance compared with one another. In order to examine the properties of these seven techniques we performed a series of similarity and classification experiments on eleven DNA microarray datasets. Our results show that in general, each feature selection technique will create diverse feature subsets when compared to the other members of the family. However when we look at classification we find that, with one exception, the techniques will produce good classification results and that the techniques will have similar performances to each other. Our recommendation, is to use the rankers Signal-to-Noise and SAM for the best classification results and to avoid Fold Change Ratio as it is consistently the worst performer of the seven rankers.","PeriodicalId":157399,"journal":{"name":"2012 11th International Conference on Machine Learning and Applications","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130530193","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}
Sriraam Natarajan, Saket Joshi, B. Saha, A. Edwards, Tushar Khot, Elizabeth Moody, K. Kersting, C. Whitlow, J. Maldjian
Magnetic resonance imaging (MRI) has emerged as an important tool to identify intermediate biomarkers of Alzheimer's disease (AD) due to its ability to measure regional changes in the brain that are thought to reflect disease severity and progression. In this paper, we set out a novel pipeline that uses volumetric MRI data collected from different subjects as input and classifies them into one of three classes: AD, mild cognitive impairment (MCI) and cognitively normal (CN). Our pipeline consists of three stages - (1) a segmentation layer where brain MRI data is divided into clinically relevant regions, (2) a classification layer that uses relational learning algorithms to make pair wise predictions between the three classes, and (3) a combination layer that combines the results of the different classes to obtain the final classification. One of the key features of our proposed approach is that it allows for domain expert's knowledge to guide the learning in all the layers. We evaluate our pipeline on 397 patients acquired from the Alzheimer's Disease Neuroimaging Initiative and demonstrate that it obtains state-of the-art performance with minimal feature engineering.
{"title":"A Machine Learning Pipeline for Three-Way Classification of Alzheimer Patients from Structural Magnetic Resonance Images of the Brain","authors":"Sriraam Natarajan, Saket Joshi, B. Saha, A. Edwards, Tushar Khot, Elizabeth Moody, K. Kersting, C. Whitlow, J. Maldjian","doi":"10.1109/ICMLA.2012.42","DOIUrl":"https://doi.org/10.1109/ICMLA.2012.42","url":null,"abstract":"Magnetic resonance imaging (MRI) has emerged as an important tool to identify intermediate biomarkers of Alzheimer's disease (AD) due to its ability to measure regional changes in the brain that are thought to reflect disease severity and progression. In this paper, we set out a novel pipeline that uses volumetric MRI data collected from different subjects as input and classifies them into one of three classes: AD, mild cognitive impairment (MCI) and cognitively normal (CN). Our pipeline consists of three stages - (1) a segmentation layer where brain MRI data is divided into clinically relevant regions, (2) a classification layer that uses relational learning algorithms to make pair wise predictions between the three classes, and (3) a combination layer that combines the results of the different classes to obtain the final classification. One of the key features of our proposed approach is that it allows for domain expert's knowledge to guide the learning in all the layers. We evaluate our pipeline on 397 patients acquired from the Alzheimer's Disease Neuroimaging Initiative and demonstrate that it obtains state-of the-art performance with minimal feature engineering.","PeriodicalId":157399,"journal":{"name":"2012 11th International Conference on Machine Learning and Applications","volume":"225 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130768548","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}
While finding natural clusters in high dimensional data is in itself a challenge, the dynamic nature of data adds another greater challenge. Many applications such as Data Warehouses and WWW demand the presence of efficient incremental clustering algorithms to handle their dynamic data. So far, numerous useful incremental clustering algorithms have been developed for large datasets such as incremental K-means, incremental DBSCAN, similarity histogram-based clustering (SHC) and mean shift. However, targeting clusters of different shapes and densities is yet to be efficiently tackled. In this work, an efficient incremental clustering algorithm (Incremental Mitosis) is proposed. It is based on Mitosis clustering algorithm which maximizes the relatedness of distances between patterns of the same cluster. The proposed algorithm is able to discover clusters of arbitrary shapes and densities in dynamic high dimensional data. Experimental results show that the proposed algorithm efficiently clusters the data and maintains the accuracy of Mitosis algorithm.
{"title":"Incremental Mitosis: Discovering Clusters of Arbitrary Shapes and Densities in Dynamic Data","authors":"Rania Ibrahim, N. Ahmed, N. A. Yousri, M. Ismail","doi":"10.1109/ICMLA.2012.26","DOIUrl":"https://doi.org/10.1109/ICMLA.2012.26","url":null,"abstract":"While finding natural clusters in high dimensional data is in itself a challenge, the dynamic nature of data adds another greater challenge. Many applications such as Data Warehouses and WWW demand the presence of efficient incremental clustering algorithms to handle their dynamic data. So far, numerous useful incremental clustering algorithms have been developed for large datasets such as incremental K-means, incremental DBSCAN, similarity histogram-based clustering (SHC) and mean shift. However, targeting clusters of different shapes and densities is yet to be efficiently tackled. In this work, an efficient incremental clustering algorithm (Incremental Mitosis) is proposed. It is based on Mitosis clustering algorithm which maximizes the relatedness of distances between patterns of the same cluster. The proposed algorithm is able to discover clusters of arbitrary shapes and densities in dynamic high dimensional data. Experimental results show that the proposed algorithm efficiently clusters the data and maintains the accuracy of Mitosis algorithm.","PeriodicalId":157399,"journal":{"name":"2012 11th International Conference on Machine Learning and Applications","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130808893","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}
This paper presents a model-based criterion for assessing the clustering results of spatial data, where both geometrical constraints and observation attributes are taken into account. An extra parameter is often used in the aim of controlling the importance of each characteristic. Since the values of both terms vary according to different realizations of data, it becomes essential to determine the parameter value which has a large influence on the clustering criterion value. Thus, an `upper-lower bound' technique is proposed to solve that problem caused by stochastic properties in both terms. In addition, we apply a normalization method to regularize the parameter value. The effectiveness of this approach is validated through the experimental results by using simulated reliability data.
{"title":"A Normalized Criterion of Spatial Clustering in Model-Based Framework","authors":"X. Wang, E. Grall-Maës, P. Beauseroy","doi":"10.1109/ICMLA.2012.99","DOIUrl":"https://doi.org/10.1109/ICMLA.2012.99","url":null,"abstract":"This paper presents a model-based criterion for assessing the clustering results of spatial data, where both geometrical constraints and observation attributes are taken into account. An extra parameter is often used in the aim of controlling the importance of each characteristic. Since the values of both terms vary according to different realizations of data, it becomes essential to determine the parameter value which has a large influence on the clustering criterion value. Thus, an `upper-lower bound' technique is proposed to solve that problem caused by stochastic properties in both terms. In addition, we apply a normalization method to regularize the parameter value. The effectiveness of this approach is validated through the experimental results by using simulated reliability data.","PeriodicalId":157399,"journal":{"name":"2012 11th International Conference on Machine Learning and Applications","volume":"9 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130826387","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}
Voice and multimedia communications are rapidly migrating from traditional networks to TCP/IP networks (Internet), where services are provisioned by SIP (Session Initiation Protocol). This paper proposes an on-line filter that examines the stream of incoming SIP messages and classifies them as good or bad. The classification is carried out in two stages: first a lexical analysis is performed to weed out those messages that do not belong to the language generated by the grammar defined by the SIP standard. After this first stage, a second filtering occurs which identifies messages that somehow differ - in structure or contents - from messages that were previously classified as good. While the first filter stage is straightforward, as the classification is crisp (either a messages belongs to the language or it does not), the second stage requires a more delicate handling, as it is not a sharp decision whether a message is semantically meaningful or not. The approach we followed for this step is based on using past experience on previously classified messages, i.e. a "learn-by-example" approach, which led to a classifier based on Support-Vector-Machines (SVM) to perform the required analysis of each incoming SIP message. The paper describes the overall architecture of the two-stage filter and then explores several points of the configuration-space for the SVM to determine a good configuration setting that will perform well when used to classify a large sample of SIP messages obtained from real traffic collected on a VoIP installation at our institution. Finally, the performance of the classification on additional messages collected from the same source is presented.
{"title":"On the Use of SVMs to Detect Anomalies in a Stream of SIP Messages","authors":"Raihana Ferdous, R. Cigno, A. Zorat","doi":"10.1109/ICMLA.2012.109","DOIUrl":"https://doi.org/10.1109/ICMLA.2012.109","url":null,"abstract":"Voice and multimedia communications are rapidly migrating from traditional networks to TCP/IP networks (Internet), where services are provisioned by SIP (Session Initiation Protocol). This paper proposes an on-line filter that examines the stream of incoming SIP messages and classifies them as good or bad. The classification is carried out in two stages: first a lexical analysis is performed to weed out those messages that do not belong to the language generated by the grammar defined by the SIP standard. After this first stage, a second filtering occurs which identifies messages that somehow differ - in structure or contents - from messages that were previously classified as good. While the first filter stage is straightforward, as the classification is crisp (either a messages belongs to the language or it does not), the second stage requires a more delicate handling, as it is not a sharp decision whether a message is semantically meaningful or not. The approach we followed for this step is based on using past experience on previously classified messages, i.e. a \"learn-by-example\" approach, which led to a classifier based on Support-Vector-Machines (SVM) to perform the required analysis of each incoming SIP message. The paper describes the overall architecture of the two-stage filter and then explores several points of the configuration-space for the SVM to determine a good configuration setting that will perform well when used to classify a large sample of SIP messages obtained from real traffic collected on a VoIP installation at our institution. Finally, the performance of the classification on additional messages collected from the same source is presented.","PeriodicalId":157399,"journal":{"name":"2012 11th International Conference on Machine Learning and Applications","volume":"4 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-12-12","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130966840","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：