Pub Date : 2012-09-01DOI: 10.1109/INTECH.2012.6457775
F. Bezzazi, A. Kartit, M. E. Marraki, D. Aboutajdine
Modern networks are becoming larger and more complex. This requires administrators and management tools to update their security policies and improve the deployment of their firewall. This latter represents a challenging and error-prone task, which aims to pass from one policy to another one while respecting the conditions of safety and security. Several studies have focused on the conception, optimization of security policy and conflict detection but rare are the researchers who tried to improve the firewall policy deployment. In this paper, we investigated the algorithm of type II policy and showed its vulnerability in some cases. As alternative, we provide a new algorithm for publishing political class type II. Further to its safety, our algorithm can be used even for the deployment of important size policies.
{"title":"Optimized strategy of deployment firewall policies","authors":"F. Bezzazi, A. Kartit, M. E. Marraki, D. Aboutajdine","doi":"10.1109/INTECH.2012.6457775","DOIUrl":"https://doi.org/10.1109/INTECH.2012.6457775","url":null,"abstract":"Modern networks are becoming larger and more complex. This requires administrators and management tools to update their security policies and improve the deployment of their firewall. This latter represents a challenging and error-prone task, which aims to pass from one policy to another one while respecting the conditions of safety and security. Several studies have focused on the conception, optimization of security policy and conflict detection but rare are the researchers who tried to improve the firewall policy deployment. In this paper, we investigated the algorithm of type II policy and showed its vulnerability in some cases. As alternative, we provide a new algorithm for publishing political class type II. Further to its safety, our algorithm can be used even for the deployment of important size policies.","PeriodicalId":369113,"journal":{"name":"Second International Conference on the Innovative Computing Technology (INTECH 2012)","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126115352","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 : 2012-09-01DOI: 10.1109/INTECH.2012.6457764
M. Fathy, E. Hagras, M. S. E. Mahallawy, M. A. Dahab
The performance of Interleave Division Multiple Access (IDMA) with a new set of low rate convolution codes is evaluated on Additive White Gaussian Noise (AWGN) channel. Despite the excessive multiuser interference, very good performance close to theoretical limit is demonstrated. The effect of varying number of users, code rate and channel characteristics on the performance is investigated. The proposed scheme offers significant complexity reduction when compared to the ones available in literature. In this work, we propose an unequal error protection code for IDMA (UEP-CIDMA) and compare its performance with the coded and uncoded IDMA system under various data bit streams. It was shown by simulations that the proposed code provides a tradeoff between cost and complexity compared with coded IDMA being the upper and lower bounds respectively, in addition to simulation for various data bit stream of convolutional encoder with different rates are 1/2 and 2/3.
{"title":"Unequal error protection coded Interleave Division Multiple Access","authors":"M. Fathy, E. Hagras, M. S. E. Mahallawy, M. A. Dahab","doi":"10.1109/INTECH.2012.6457764","DOIUrl":"https://doi.org/10.1109/INTECH.2012.6457764","url":null,"abstract":"The performance of Interleave Division Multiple Access (IDMA) with a new set of low rate convolution codes is evaluated on Additive White Gaussian Noise (AWGN) channel. Despite the excessive multiuser interference, very good performance close to theoretical limit is demonstrated. The effect of varying number of users, code rate and channel characteristics on the performance is investigated. The proposed scheme offers significant complexity reduction when compared to the ones available in literature. In this work, we propose an unequal error protection code for IDMA (UEP-CIDMA) and compare its performance with the coded and uncoded IDMA system under various data bit streams. It was shown by simulations that the proposed code provides a tradeoff between cost and complexity compared with coded IDMA being the upper and lower bounds respectively, in addition to simulation for various data bit stream of convolutional encoder with different rates are 1/2 and 2/3.","PeriodicalId":369113,"journal":{"name":"Second International Conference on the Innovative Computing Technology (INTECH 2012)","volume":"126 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128070136","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 : 2012-09-01DOI: 10.1109/INTECH.2012.6457807
O. Baida, N. Hamzaoui, A. Sedqui, A. Lyhyaoui
Recommender system is a system that helps users find interesting items. Actually, collaborative filtering technology is one of the most successful techniques in recommender system. In this article we propose a new approach based on the rating of the users that is similar to the active one. In the literature, we find a lot of approaches able to recommend items to the user. Aiming to offer a list of interesting items, we use a hybrid approach of collaborative filtering that performs better than others. Our collaborative filtering approach is based on the graph theory, so we use the dissimilarity matrix as a spanning tree with minimum weight based on Kruskal algorithm. We define a group of criteria that help to determine the best items to recommend without computing the rating prediction.
{"title":"Recommendation based on co-similarity and spanning tree with minimum weight","authors":"O. Baida, N. Hamzaoui, A. Sedqui, A. Lyhyaoui","doi":"10.1109/INTECH.2012.6457807","DOIUrl":"https://doi.org/10.1109/INTECH.2012.6457807","url":null,"abstract":"Recommender system is a system that helps users find interesting items. Actually, collaborative filtering technology is one of the most successful techniques in recommender system. In this article we propose a new approach based on the rating of the users that is similar to the active one. In the literature, we find a lot of approaches able to recommend items to the user. Aiming to offer a list of interesting items, we use a hybrid approach of collaborative filtering that performs better than others. Our collaborative filtering approach is based on the graph theory, so we use the dissimilarity matrix as a spanning tree with minimum weight based on Kruskal algorithm. We define a group of criteria that help to determine the best items to recommend without computing the rating prediction.","PeriodicalId":369113,"journal":{"name":"Second International Conference on the Innovative Computing Technology (INTECH 2012)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128448584","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 : 2012-09-01DOI: 10.1109/INTECH.2012.6457768
M. El Boujnouni, M. Jedra, N. Zahid
In conventional support vector data description (SVDD), for each class we look for the smallest sphere that encloses its data. in the decision phase a sample is classified into class i only when the value of the ith decision function is positive. following this architecture, an unclassifiable region (s) can be appeared if the values of more than one decision function are positives. To overcome this problem, we propose a new simple and powerful decision function, which is used only in the overlappeds regions, this membership function can be calculated in the feature space and can be represented by kernels functions. This method gives good performance on reducing the effects of overlap and significantly improves the classification. We demonstrate the performance of our decision function using six benchmark datasets.
{"title":"New decision function for support vector data description","authors":"M. El Boujnouni, M. Jedra, N. Zahid","doi":"10.1109/INTECH.2012.6457768","DOIUrl":"https://doi.org/10.1109/INTECH.2012.6457768","url":null,"abstract":"In conventional support vector data description (SVDD), for each class we look for the smallest sphere that encloses its data. in the decision phase a sample is classified into class i only when the value of the ith decision function is positive. following this architecture, an unclassifiable region (s) can be appeared if the values of more than one decision function are positives. To overcome this problem, we propose a new simple and powerful decision function, which is used only in the overlappeds regions, this membership function can be calculated in the feature space and can be represented by kernels functions. This method gives good performance on reducing the effects of overlap and significantly improves the classification. We demonstrate the performance of our decision function using six benchmark datasets.","PeriodicalId":369113,"journal":{"name":"Second International Conference on the Innovative Computing Technology (INTECH 2012)","volume":"67 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127246090","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 : 2012-09-01DOI: 10.1109/INTECH.2012.6457766
M. F. Ali, A. Barnawi, A. Bashar
This paper presents a novel evaluation study of the Cloud Computing technology, with a focused emphasis on the Cloud Storage mechanisms and the way they are affecting the progress of the present Cloud Services. Considering the exponential growth of the user data and its impact on the Cloud Storage infrastructure, this work provides two major contributions through comprehensive performance evaluations. Firstly, it proposes a unique 10-point performance evaluation framework for existing Cloud Storage infrastructure and applies it for evaluating six major Cloud Storage Service Providers currently in the market. Secondly, it presents a detailed insightful assessment of eighteen most popular Cloud Storage Hardware vendors with respect to the storage technologies being implemented by them. In conclusion, it takes stock of the current trends on optimizing storage infrastructure for Cloud Computing and predicting future research possibilities in this rapidly growing technology.
{"title":"Performance analysis framework to optimize storage infrastructure for Cloud Computing","authors":"M. F. Ali, A. Barnawi, A. Bashar","doi":"10.1109/INTECH.2012.6457766","DOIUrl":"https://doi.org/10.1109/INTECH.2012.6457766","url":null,"abstract":"This paper presents a novel evaluation study of the Cloud Computing technology, with a focused emphasis on the Cloud Storage mechanisms and the way they are affecting the progress of the present Cloud Services. Considering the exponential growth of the user data and its impact on the Cloud Storage infrastructure, this work provides two major contributions through comprehensive performance evaluations. Firstly, it proposes a unique 10-point performance evaluation framework for existing Cloud Storage infrastructure and applies it for evaluating six major Cloud Storage Service Providers currently in the market. Secondly, it presents a detailed insightful assessment of eighteen most popular Cloud Storage Hardware vendors with respect to the storage technologies being implemented by them. In conclusion, it takes stock of the current trends on optimizing storage infrastructure for Cloud Computing and predicting future research possibilities in this rapidly growing technology.","PeriodicalId":369113,"journal":{"name":"Second International Conference on the Innovative Computing Technology (INTECH 2012)","volume":"2020 4","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114040478","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 : 2012-09-01DOI: 10.1109/INTECH.2012.6457770
A. Smaili, A. Lasfar, M. Sbihi
During the last decade, the technique of optical character recognition has a particular research interest that laboratories continue to improve and develop. This technique is a good alternative to entering text, the availability of various tools involved in the singularity of this service. Undoubtedly, the development of such tools will provide a considerable gain in time and costs. The basis of the Arabic language, prompted us to work hard for a successful control of its optical character recognition, otherwise it is necessary to develop another approaches, therefore an approach based on finding a word-image in a text-image. Finally, the objective of this work is to conduct research in a word-image-text image of the Moroccan official bulletin. Finally, the objective of this work is to conduct research a wordimage in text image of the Moroccan official bulletin.
{"title":"A word-image's system searching: Case of printed Arabic word","authors":"A. Smaili, A. Lasfar, M. Sbihi","doi":"10.1109/INTECH.2012.6457770","DOIUrl":"https://doi.org/10.1109/INTECH.2012.6457770","url":null,"abstract":"During the last decade, the technique of optical character recognition has a particular research interest that laboratories continue to improve and develop. This technique is a good alternative to entering text, the availability of various tools involved in the singularity of this service. Undoubtedly, the development of such tools will provide a considerable gain in time and costs. The basis of the Arabic language, prompted us to work hard for a successful control of its optical character recognition, otherwise it is necessary to develop another approaches, therefore an approach based on finding a word-image in a text-image. Finally, the objective of this work is to conduct research in a word-image-text image of the Moroccan official bulletin. Finally, the objective of this work is to conduct research a wordimage in text image of the Moroccan official bulletin.","PeriodicalId":369113,"journal":{"name":"Second International Conference on the Innovative Computing Technology (INTECH 2012)","volume":"336 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116261447","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 : 2012-09-01DOI: 10.1109/INTECH.2012.6457776
Mohammad Kia, Shabnam Kia, Neda Davoudi, Ramyar Biniazan
It is inconvenient for housekeeper parents to constantly watch over their newborn baby while doing their work or chores. This paper proposes a simple voice recognition system which can be applied practically for designing a device with capability to detect a baby's cry and informing the parents automatically. There are a lot of similar projects and experiments which have been performed recently, but most of them are about recognizing and classifying different types of crying (like for hunger, etc.) and have used complex methods of implementation such as neural network. But in this paper our aim is to merely detect infant's crying and our solution is to use a fuzzy classifier which is easy to implement. The overall algorithm is to evaluate the resemblance of the infant's voice signal with the data stored in a database, which is already prepared by recording some cry and laughter samples, using an automatic fuzzy classifier system which can lead to detection of cry or laughter. This algorithm can serve as a reliable foundation on which the future creation of a portable real-time, automatic voice detection system device can be based. It is a pretty formidable task to implement complex algorithms, such as neural networks, on common available microcontrollers, however we proposed much simpler procedure which enables us to develop a low cost device. To achieve our aim, we have created a database of sample cry and laughter signals and written a sample Matlab program for carrying out the real-time frequency-domain calculations and a sample visual program in Labview programming environment for interfacing with user.
{"title":"A detection system of infant cry using fuzzy classification including dialing alarm calls function","authors":"Mohammad Kia, Shabnam Kia, Neda Davoudi, Ramyar Biniazan","doi":"10.1109/INTECH.2012.6457776","DOIUrl":"https://doi.org/10.1109/INTECH.2012.6457776","url":null,"abstract":"It is inconvenient for housekeeper parents to constantly watch over their newborn baby while doing their work or chores. This paper proposes a simple voice recognition system which can be applied practically for designing a device with capability to detect a baby's cry and informing the parents automatically. There are a lot of similar projects and experiments which have been performed recently, but most of them are about recognizing and classifying different types of crying (like for hunger, etc.) and have used complex methods of implementation such as neural network. But in this paper our aim is to merely detect infant's crying and our solution is to use a fuzzy classifier which is easy to implement. The overall algorithm is to evaluate the resemblance of the infant's voice signal with the data stored in a database, which is already prepared by recording some cry and laughter samples, using an automatic fuzzy classifier system which can lead to detection of cry or laughter. This algorithm can serve as a reliable foundation on which the future creation of a portable real-time, automatic voice detection system device can be based. It is a pretty formidable task to implement complex algorithms, such as neural networks, on common available microcontrollers, however we proposed much simpler procedure which enables us to develop a low cost device. To achieve our aim, we have created a database of sample cry and laughter signals and written a sample Matlab program for carrying out the real-time frequency-domain calculations and a sample visual program in Labview programming environment for interfacing with user.","PeriodicalId":369113,"journal":{"name":"Second International Conference on the Innovative Computing Technology (INTECH 2012)","volume":"27 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114623284","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 : 2012-09-01DOI: 10.1109/INTECH.2012.6457810
A. E. Bouchti, A. Haqiq
With the rise of cyber-attack activities in the recent years, research in this area has gained immense emphasis. One of such research efforts is modeling of cyber-attacks. In this context, several modeling approaches have been developed, such as approaches based on attack trees (AT). In this paper, we propose Colored Petri Net (CoPNet) modeling approach by extending the attack trees with new modeling constructs and analysis approaches. CoPNet based attack model is flexible enough to model Internet intrusion, including the static and dynamic features of the intrusion. The process and rules of building CoPNet based attack model from AT are also presented. In order to evaluate the risk of intrusion, some cost elements are added to CoPNet based attack modeling. We show how attack trees can be converted and analyzed in CoPNets. Finally, we provide a case study that illustrates the CoPNet approach.
{"title":"Cyber-attack for BGP systems using CoPNet model","authors":"A. E. Bouchti, A. Haqiq","doi":"10.1109/INTECH.2012.6457810","DOIUrl":"https://doi.org/10.1109/INTECH.2012.6457810","url":null,"abstract":"With the rise of cyber-attack activities in the recent years, research in this area has gained immense emphasis. One of such research efforts is modeling of cyber-attacks. In this context, several modeling approaches have been developed, such as approaches based on attack trees (AT). In this paper, we propose Colored Petri Net (CoPNet) modeling approach by extending the attack trees with new modeling constructs and analysis approaches. CoPNet based attack model is flexible enough to model Internet intrusion, including the static and dynamic features of the intrusion. The process and rules of building CoPNet based attack model from AT are also presented. In order to evaluate the risk of intrusion, some cost elements are added to CoPNet based attack modeling. We show how attack trees can be converted and analyzed in CoPNets. Finally, we provide a case study that illustrates the CoPNet approach.","PeriodicalId":369113,"journal":{"name":"Second International Conference on the Innovative Computing Technology (INTECH 2012)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126843997","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 : 2012-09-01DOI: 10.1109/INTECH.2012.6457761
S. Abdellaoui, Y. Fakhri, D. Aboutajdine
Energy is an extremely limited resource in wireless sensor networks (WSNs). Thereby, various power management strategies have been proposed to reduce the quantity of consumed energy. In this article, the aim of our approach is to provide the optimal transmission power in order to maximize the network lifetime using the orthogonal channels and taking into account a Symbol Error Rate (SER) constraint at the Fusion Center (FC). In order to analyze our approach, we will first study the case where the information passes by a single relay to reach the destination creating several boughs. We call this case “virtual mimo with a single relay”. Second, we will consider a specific case with a single bough and several relays and then generalize a global case with several relays and several boughs. In addition, we suppose that a Maximum Ratio Combining (MRC) is used by the receiver as a detector, and Amplify and Forward (AF) is used as a relaying strategy where the channel gain coefficients are known. Simulation results show the viability of our approach.
{"title":"Optimum power allocation for Amplify-and-Forward cooperation strategy in WSN","authors":"S. Abdellaoui, Y. Fakhri, D. Aboutajdine","doi":"10.1109/INTECH.2012.6457761","DOIUrl":"https://doi.org/10.1109/INTECH.2012.6457761","url":null,"abstract":"Energy is an extremely limited resource in wireless sensor networks (WSNs). Thereby, various power management strategies have been proposed to reduce the quantity of consumed energy. In this article, the aim of our approach is to provide the optimal transmission power in order to maximize the network lifetime using the orthogonal channels and taking into account a Symbol Error Rate (SER) constraint at the Fusion Center (FC). In order to analyze our approach, we will first study the case where the information passes by a single relay to reach the destination creating several boughs. We call this case “virtual mimo with a single relay”. Second, we will consider a specific case with a single bough and several relays and then generalize a global case with several relays and several boughs. In addition, we suppose that a Maximum Ratio Combining (MRC) is used by the receiver as a detector, and Amplify and Forward (AF) is used as a relaying strategy where the channel gain coefficients are known. Simulation results show the viability of our approach.","PeriodicalId":369113,"journal":{"name":"Second International Conference on the Innovative Computing Technology (INTECH 2012)","volume":"84 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132708142","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 : 2012-09-01DOI: 10.1109/INTECH.2012.6457778
Ikram Chairi, Souad Alaoui, A. Lyhyaoui
The majority of learning systems usually assume that training sets are balanced, however, in real world data this hypothesis is not always true. The problem of between-class imbalance is a challenge that has attracted growing attention from both academia and industry, because of its critical influence on the performance of learning systems. Many solutions were proposed to resolve this problem: Generally, the common practice for dealing with imbalanced data sets is to rebalance them artificially by using sampling methods. In this paper, we propose a method based on Sample Selection (SS), to deal with the problem of between class imbalance. We consider that creating balance between classes by paying more attention to those examples located near the border line improves the performance of the classifier. To reduce the computational cost of selecting samples, we propose a clustering method as a first step in order to determine the critical centers, and then we select samples from those critical clusters. Experimental results with Multi-Layer Perceptron (MLP) architecture, on well known Intrusion Detection data set, support the usefulness of our approach.
{"title":"Intrusion Detection based Sample Selection for imbalanced data distribution","authors":"Ikram Chairi, Souad Alaoui, A. Lyhyaoui","doi":"10.1109/INTECH.2012.6457778","DOIUrl":"https://doi.org/10.1109/INTECH.2012.6457778","url":null,"abstract":"The majority of learning systems usually assume that training sets are balanced, however, in real world data this hypothesis is not always true. The problem of between-class imbalance is a challenge that has attracted growing attention from both academia and industry, because of its critical influence on the performance of learning systems. Many solutions were proposed to resolve this problem: Generally, the common practice for dealing with imbalanced data sets is to rebalance them artificially by using sampling methods. In this paper, we propose a method based on Sample Selection (SS), to deal with the problem of between class imbalance. We consider that creating balance between classes by paying more attention to those examples located near the border line improves the performance of the classifier. To reduce the computational cost of selecting samples, we propose a clustering method as a first step in order to determine the critical centers, and then we select samples from those critical clusters. Experimental results with Multi-Layer Perceptron (MLP) architecture, on well known Intrusion Detection data set, support the usefulness of our approach.","PeriodicalId":369113,"journal":{"name":"Second International Conference on the Innovative Computing Technology (INTECH 2012)","volume":"103 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2012-09-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134344399","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：