Pub Date : 2016-12-01DOI: 10.1109/SSCI.2016.7850248
Ibrahim Ozturk, D. Halliday
In this paper, a simple structure of two-layer feed-forward spiking neural network (SNN) is developed which is trained by reward-modulated Spike Timing Dependent Plasticity (STDP). Neurons based on leaky integrate-and-fire (LIF) neuron model are trained to associate input temporal sequences with a desired output spike pattern, both consisting of multiple spikes. A biologically plausible Reward-Modulated STDP learning rule is used so that the network can efficiently converge optimal spike generation. The relative timing of pre- and postsynaptic firings can only modify synaptic weights once the reward has occurred. The history of Hebbian events are stored in the synaptic eligibility traces. STDP process are applied to all synapses with different delays. We experimentally demonstrate a benchmark with spatio-temporally encoded spike pairs. Results demonstrate successful transformations with high accuracy and quick convergence during learning cycles. Therefore, the proposed SNN architecture with modulated STDP can learn how to map temporally encoded spike trains based on Poisson processes in a stable manner.
{"title":"Mapping spatio-temporally encoded patterns by reward-modulated STDP in Spiking neurons","authors":"Ibrahim Ozturk, D. Halliday","doi":"10.1109/SSCI.2016.7850248","DOIUrl":"https://doi.org/10.1109/SSCI.2016.7850248","url":null,"abstract":"In this paper, a simple structure of two-layer feed-forward spiking neural network (SNN) is developed which is trained by reward-modulated Spike Timing Dependent Plasticity (STDP). Neurons based on leaky integrate-and-fire (LIF) neuron model are trained to associate input temporal sequences with a desired output spike pattern, both consisting of multiple spikes. A biologically plausible Reward-Modulated STDP learning rule is used so that the network can efficiently converge optimal spike generation. The relative timing of pre- and postsynaptic firings can only modify synaptic weights once the reward has occurred. The history of Hebbian events are stored in the synaptic eligibility traces. STDP process are applied to all synapses with different delays. We experimentally demonstrate a benchmark with spatio-temporally encoded spike pairs. Results demonstrate successful transformations with high accuracy and quick convergence during learning cycles. Therefore, the proposed SNN architecture with modulated STDP can learn how to map temporally encoded spike trains based on Poisson processes in a stable manner.","PeriodicalId":120288,"journal":{"name":"2016 IEEE Symposium Series on Computational Intelligence (SSCI)","volume":"74 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126263922","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 : 2016-12-01DOI: 10.1109/SSCI.2016.7850135
S. Hiwa, Yuuki Kohri, Keisuke Hachisuka, T. Hiroyasu
Functional connectivity, which is indicated by time-course correlations of brain activities among different brain regions, is one of the most useful metrics to represent human brain states. In functional connectivity analysis (FCA), the whole brain is parcellated into a certain number of regions based on anatomical atlases, and the mean time series of brain activities are calculated. Then, the correlation between mean signals of two regions is repeatedly calculated for all combinations of regions, and finally, we obtain the correlation matrix of the whole brain. FCA allows us to understand which regions activate cooperatively during specific stimulus or tasks. In this study, we attempt to represent human brain states using functional connectivity as feature vectors. As there are a number of brain regions, it is difficult to determine which regions are prominent to represent the brain state. Therefore, we proposed an automatic region-of-interest (ROI) extraction method to classify human brain states. Time-series brain activities were measured by functional magnetic resonance imaging (fMRI), and FCA was performed. Each element of the correlation matrix was used as a feature vector for brain state classification, and element characteristics were learned using supervised learning methods. The elements used as feature vectors, i.e., ROIs, were determined automatically using a genetic algorithm to maximize the classification accuracy of brain states. fMRI data measured during two emotional conditions, i.e., pleasant and unpleasant emotions, were used to show the effectiveness of the proposed method. Numerical experiments revealed that the proposed method could extract the superior frontal gyrus, orbitofrontal cortex, cuneus, cerebellum, and cerebellar vermis as ROIs associated with pleasant and unpleasant emotions.
{"title":"Region-of-interest extraction of fMRI data using genetic algorithms","authors":"S. Hiwa, Yuuki Kohri, Keisuke Hachisuka, T. Hiroyasu","doi":"10.1109/SSCI.2016.7850135","DOIUrl":"https://doi.org/10.1109/SSCI.2016.7850135","url":null,"abstract":"Functional connectivity, which is indicated by time-course correlations of brain activities among different brain regions, is one of the most useful metrics to represent human brain states. In functional connectivity analysis (FCA), the whole brain is parcellated into a certain number of regions based on anatomical atlases, and the mean time series of brain activities are calculated. Then, the correlation between mean signals of two regions is repeatedly calculated for all combinations of regions, and finally, we obtain the correlation matrix of the whole brain. FCA allows us to understand which regions activate cooperatively during specific stimulus or tasks. In this study, we attempt to represent human brain states using functional connectivity as feature vectors. As there are a number of brain regions, it is difficult to determine which regions are prominent to represent the brain state. Therefore, we proposed an automatic region-of-interest (ROI) extraction method to classify human brain states. Time-series brain activities were measured by functional magnetic resonance imaging (fMRI), and FCA was performed. Each element of the correlation matrix was used as a feature vector for brain state classification, and element characteristics were learned using supervised learning methods. The elements used as feature vectors, i.e., ROIs, were determined automatically using a genetic algorithm to maximize the classification accuracy of brain states. fMRI data measured during two emotional conditions, i.e., pleasant and unpleasant emotions, were used to show the effectiveness of the proposed method. Numerical experiments revealed that the proposed method could extract the superior frontal gyrus, orbitofrontal cortex, cuneus, cerebellum, and cerebellar vermis as ROIs associated with pleasant and unpleasant emotions.","PeriodicalId":120288,"journal":{"name":"2016 IEEE Symposium Series on Computational Intelligence (SSCI)","volume":"7 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125709756","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 : 2016-12-01DOI: 10.1109/SSCI.2016.7850202
Martin Hyrs, J. Schwarz
Estimation of distribution algorithms (EDAs) are stochastic optimization techniques that are based on building and sampling a probability model. Copula theory provides methods that simplify the estimation of the probability model. To improve the efficiency of current copula based EDAs (CEDAs) new modifications of parallel CEDA were proposed. We investigated eight variants of island-based algorithms utilizing the capability of promising copula families, inter-island migration and additional adaptation of marginal parameters using CT-AVS technique. The proposed algorithms were tested on two sets of well-known standard optimization benchmarks in the continuous domain. The results of the experiments validate the efficiency of our algorithms.
{"title":"Advanced parallel copula based EDA","authors":"Martin Hyrs, J. Schwarz","doi":"10.1109/SSCI.2016.7850202","DOIUrl":"https://doi.org/10.1109/SSCI.2016.7850202","url":null,"abstract":"Estimation of distribution algorithms (EDAs) are stochastic optimization techniques that are based on building and sampling a probability model. Copula theory provides methods that simplify the estimation of the probability model. To improve the efficiency of current copula based EDAs (CEDAs) new modifications of parallel CEDA were proposed. We investigated eight variants of island-based algorithms utilizing the capability of promising copula families, inter-island migration and additional adaptation of marginal parameters using CT-AVS technique. The proposed algorithms were tested on two sets of well-known standard optimization benchmarks in the continuous domain. The results of the experiments validate the efficiency of our algorithms.","PeriodicalId":120288,"journal":{"name":"2016 IEEE Symposium Series on Computational Intelligence (SSCI)","volume":"19 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121684528","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 : 2016-12-01DOI: 10.1109/SSCI.2016.7850183
B. Mocialov, P. A. Vargas, M. Couceiro
This paper presents preliminary investigations on the evolution of indirect communication between two agents. In the future, behaviours of robots in the RoboCup1 competition should resemble the behaviours of the human players. One common trait of this behaviour is the indirect communication. Within the human-robot-interaction, indirect communication can either be the principal or supporting method for information exchange. This paper summarises previous work on the topic and presents the design of a self-organised system for gesture recognition. Although, preliminary results show that the proposed system requires further feature extraction improvements and evaluations on various public datasets, the system is capable of performing classification of gestures. Further research is required to fully investigate potential extensions to the system that would be able to support real indirect communication in human-robot interaction scenarios.
{"title":"Towards the evolution of indirect communication for social robots","authors":"B. Mocialov, P. A. Vargas, M. Couceiro","doi":"10.1109/SSCI.2016.7850183","DOIUrl":"https://doi.org/10.1109/SSCI.2016.7850183","url":null,"abstract":"This paper presents preliminary investigations on the evolution of indirect communication between two agents. In the future, behaviours of robots in the RoboCup1 competition should resemble the behaviours of the human players. One common trait of this behaviour is the indirect communication. Within the human-robot-interaction, indirect communication can either be the principal or supporting method for information exchange. This paper summarises previous work on the topic and presents the design of a self-organised system for gesture recognition. Although, preliminary results show that the proposed system requires further feature extraction improvements and evaluations on various public datasets, the system is capable of performing classification of gestures. Further research is required to fully investigate potential extensions to the system that would be able to support real indirect communication in human-robot interaction scenarios.","PeriodicalId":120288,"journal":{"name":"2016 IEEE Symposium Series on Computational Intelligence (SSCI)","volume":"146 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122917138","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 : 2016-12-01DOI: 10.1109/SSCI.2016.7849908
Neal Wagner, C. Sahin, M. Winterrose, J. Riordan, Jaime Peña, D. Hanson, W. Streilein
Network segmentation is a security measure that partitions a network into sections or segments to restrict the movement of a cyber attacker and make it difficult for her to gain access to valuable network resources. This threat-mitigating practice has been recommended by several information security agencies. While it is clear that segmentation is a critical defensive mitigation against cyber threats, it is not clear how to properly apply it. Current standards only offer vague guidance on how to apply segmentation and, thus, practitioners must rely on judgment. This paper examines the problem from a decision support perspective: that is, how can an appropriate segmentation for a given network environment be selected? We propose a novel method for supporting such a decision that utilizes an approach based on heuristic search and agent-based simulation. We have implemented a first prototype of our method and illustrate its use via a case study on a representative network environment.
{"title":"Towards automated cyber decision support: A case study on network segmentation for security","authors":"Neal Wagner, C. Sahin, M. Winterrose, J. Riordan, Jaime Peña, D. Hanson, W. Streilein","doi":"10.1109/SSCI.2016.7849908","DOIUrl":"https://doi.org/10.1109/SSCI.2016.7849908","url":null,"abstract":"Network segmentation is a security measure that partitions a network into sections or segments to restrict the movement of a cyber attacker and make it difficult for her to gain access to valuable network resources. This threat-mitigating practice has been recommended by several information security agencies. While it is clear that segmentation is a critical defensive mitigation against cyber threats, it is not clear how to properly apply it. Current standards only offer vague guidance on how to apply segmentation and, thus, practitioners must rely on judgment. This paper examines the problem from a decision support perspective: that is, how can an appropriate segmentation for a given network environment be selected? We propose a novel method for supporting such a decision that utilizes an approach based on heuristic search and agent-based simulation. We have implemented a first prototype of our method and illustrate its use via a case study on a representative network environment.","PeriodicalId":120288,"journal":{"name":"2016 IEEE Symposium Series on Computational Intelligence (SSCI)","volume":"142 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"131756587","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 : 2016-12-01DOI: 10.1109/SSCI.2016.7849954
Abeer Alzubaidi, David J. Brown, G. Cosma, A. Pockley
Reducing the number of features whilst maintaining an acceptable classification accuracy is a fundamental step in the process of constructing cancer predictive models. In this work, we introduce a novel hybrid (MI-LDA) feature selection approach for the diagnosis of ovarian cancer. This hybrid approach is embedded within a global optimization framework and offers a promising improvement on feature selection and classification accuracy processes. Global Mutual Information (MI) based feature selection optimizes the search process of finding best feature subsets in order to select the highly correlated predictors for ovarian cancer diagnosis. The maximal discriminative cancer predictors are then passed to a Linear Discriminant Analysis (LDA) classifier, and a Genetic Algorithm (GA) is applied to optimise the search process with respect to the estimated error rate of the LDA classifier (MI-LDA). Experiments were performed using an ovarian cancer dataset obtained from the FDA-NCI Clinical Proteomics Program Databank. The performance of the hybrid feature selection approach was evaluated using the Support Vector Machine (SVM) classifier and the LDA classifier. A comparison of the results revealed that the proposed (MI-LDA)-LDA model outperformed the (MI-LDA)-SVM model on selecting the maximal discriminative feature subset and achieved the highest predictive accuracy. The proposed system can therefore be used as an efficient tool for finding predictors and patterns in serum (blood)-derived proteomic data for the detection of ovarian cancer.
{"title":"A new hybrid global optimization approach for selecting clinical and biological features that are relevant to the effective diagnosis of ovarian cancer","authors":"Abeer Alzubaidi, David J. Brown, G. Cosma, A. Pockley","doi":"10.1109/SSCI.2016.7849954","DOIUrl":"https://doi.org/10.1109/SSCI.2016.7849954","url":null,"abstract":"Reducing the number of features whilst maintaining an acceptable classification accuracy is a fundamental step in the process of constructing cancer predictive models. In this work, we introduce a novel hybrid (MI-LDA) feature selection approach for the diagnosis of ovarian cancer. This hybrid approach is embedded within a global optimization framework and offers a promising improvement on feature selection and classification accuracy processes. Global Mutual Information (MI) based feature selection optimizes the search process of finding best feature subsets in order to select the highly correlated predictors for ovarian cancer diagnosis. The maximal discriminative cancer predictors are then passed to a Linear Discriminant Analysis (LDA) classifier, and a Genetic Algorithm (GA) is applied to optimise the search process with respect to the estimated error rate of the LDA classifier (MI-LDA). Experiments were performed using an ovarian cancer dataset obtained from the FDA-NCI Clinical Proteomics Program Databank. The performance of the hybrid feature selection approach was evaluated using the Support Vector Machine (SVM) classifier and the LDA classifier. A comparison of the results revealed that the proposed (MI-LDA)-LDA model outperformed the (MI-LDA)-SVM model on selecting the maximal discriminative feature subset and achieved the highest predictive accuracy. The proposed system can therefore be used as an efficient tool for finding predictors and patterns in serum (blood)-derived proteomic data for the detection of ovarian cancer.","PeriodicalId":120288,"journal":{"name":"2016 IEEE Symposium Series on Computational Intelligence (SSCI)","volume":"473 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127008676","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 : 2016-12-01DOI: 10.1109/SSCI.2016.7849866
Pierfrancesco Cervellini, A. G. Menezes, Vijay Mago
The search for Trendsetters in social networks turned to be a complex research topic that has gained much attention. The work here presented uses big data analytics to find who better spreads the word in a social network and is innovative in their choices. The analysis on the Yelp platform can be divided in three parts: first, we justify the use of Tips frequency as a variable to profile business popularity. Second we analyze Tips frequency to select businesses that fit a growing popularity profile. And third we graph mine the sociographs generated by the users that interacted with each selected business. Top nodes are ranked by using Indegree, Eigenvector centrality, Pagerank and a Trendsetter algorithms, and we compare the relative performance of each algorithm. Our findings indicate that the Trendsetter ranking algorithm is the most performant at finding nodes that best reflect the Trendsetter properties.
{"title":"Finding Trendsetters on Yelp Dataset","authors":"Pierfrancesco Cervellini, A. G. Menezes, Vijay Mago","doi":"10.1109/SSCI.2016.7849866","DOIUrl":"https://doi.org/10.1109/SSCI.2016.7849866","url":null,"abstract":"The search for Trendsetters in social networks turned to be a complex research topic that has gained much attention. The work here presented uses big data analytics to find who better spreads the word in a social network and is innovative in their choices. The analysis on the Yelp platform can be divided in three parts: first, we justify the use of Tips frequency as a variable to profile business popularity. Second we analyze Tips frequency to select businesses that fit a growing popularity profile. And third we graph mine the sociographs generated by the users that interacted with each selected business. Top nodes are ranked by using Indegree, Eigenvector centrality, Pagerank and a Trendsetter algorithms, and we compare the relative performance of each algorithm. Our findings indicate that the Trendsetter ranking algorithm is the most performant at finding nodes that best reflect the Trendsetter properties.","PeriodicalId":120288,"journal":{"name":"2016 IEEE Symposium Series on Computational Intelligence (SSCI)","volume":"286 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"116511705","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 : 2016-12-01DOI: 10.1109/SSCI.2016.7850015
Ezequiel Avilés-Ochoa, Ernesto León-Castro, J. M. Lindahl, A. M. G. Lafuente
This paper presents the results of using experton, forgotten effects and heavy moving averages operators in three traditional models based purchasing power parity (PPP) model to forecast exchange rate. Therefore, the use of these methods is to improve the forecast error under scenarios of volatility and uncertainty, such as the financial markets and more precise in exchange rate. The heavy ordered weighted moving average weighted average (HOWMAWA) operator is introduced. This new operator includes the weighted average in the usual heavy ordered weighted moving average (HOWMA) operator, considering a degree of importance for each concept that includes the operator. The use of experton and forgotten effects methodology represents the information of the experts in the field and with that information were obtained hidden variables or second degree relations. The results show that the inclusion of the forgotten effects and heavy moving average operators improve our results and reduce the forecast error.
{"title":"Forgotten effects and heavy moving averages in exchange rate forecasting","authors":"Ezequiel Avilés-Ochoa, Ernesto León-Castro, J. M. Lindahl, A. M. G. Lafuente","doi":"10.1109/SSCI.2016.7850015","DOIUrl":"https://doi.org/10.1109/SSCI.2016.7850015","url":null,"abstract":"This paper presents the results of using experton, forgotten effects and heavy moving averages operators in three traditional models based purchasing power parity (PPP) model to forecast exchange rate. Therefore, the use of these methods is to improve the forecast error under scenarios of volatility and uncertainty, such as the financial markets and more precise in exchange rate. The heavy ordered weighted moving average weighted average (HOWMAWA) operator is introduced. This new operator includes the weighted average in the usual heavy ordered weighted moving average (HOWMA) operator, considering a degree of importance for each concept that includes the operator. The use of experton and forgotten effects methodology represents the information of the experts in the field and with that information were obtained hidden variables or second degree relations. The results show that the inclusion of the forgotten effects and heavy moving average operators improve our results and reduce the forecast error.","PeriodicalId":120288,"journal":{"name":"2016 IEEE Symposium Series on Computational Intelligence (SSCI)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127826735","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 : 2016-12-01DOI: 10.1109/SSCI.2016.7850078
Duc C. Le, A. N. Zincir-Heywood, M. Heywood
Botnets represent one of the most destructive cybersecurity threats. Given the evolution of the structures and protocols botnets use, many machine learning approaches have been proposed for botnet analysis and detection. In the literature, intrusion and anomaly detection systems based on unsupervised learning techniques showed promising performances. In this paper, we investigate the capability of employing the Self-Organizing Map (SOM), an unsupervised learning technique as a data analytics system. In doing so, our aim is to understand how far such an approach could be pushed to analyze unknown traffic to detect botnets. To this end, we employed three different unsupervised training schemes using publicly available botnet data sets. Our results show that SOMs possess high potential as a data analytics tool on unknown traffic. They can identify the botnet and normal flows with high confidence approximately 99% of the time on the data sets employed in this work.
{"title":"Data analytics on network traffic flows for botnet behaviour detection","authors":"Duc C. Le, A. N. Zincir-Heywood, M. Heywood","doi":"10.1109/SSCI.2016.7850078","DOIUrl":"https://doi.org/10.1109/SSCI.2016.7850078","url":null,"abstract":"Botnets represent one of the most destructive cybersecurity threats. Given the evolution of the structures and protocols botnets use, many machine learning approaches have been proposed for botnet analysis and detection. In the literature, intrusion and anomaly detection systems based on unsupervised learning techniques showed promising performances. In this paper, we investigate the capability of employing the Self-Organizing Map (SOM), an unsupervised learning technique as a data analytics system. In doing so, our aim is to understand how far such an approach could be pushed to analyze unknown traffic to detect botnets. To this end, we employed three different unsupervised training schemes using publicly available botnet data sets. Our results show that SOMs possess high potential as a data analytics tool on unknown traffic. They can identify the botnet and normal flows with high confidence approximately 99% of the time on the data sets employed in this work.","PeriodicalId":120288,"journal":{"name":"2016 IEEE Symposium Series on Computational Intelligence (SSCI)","volume":"390 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133354505","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 : 2016-12-01DOI: 10.1109/SSCI.2016.7849938
Willem S. van Heerden, A. Engelbrecht
Self-Organizing Maps (SOMs) are unsupervised neural networks that build data models. Neuron labeling attaches descriptive textual labels to the neurons making up a SOM, and is an important component of SOM-based exploratory data analysis (EDA) and data mining (DM). Several neuron labeling approaches tend to leave some neurons unlabeled. The interaction between unlabeled neurons and SOM model accuracy affect the choice of labeling algorithm for SOM-based EDA and DM, but has not been previously investigated. This paper applies the widely used example-centric neuron labeling algorithm to several classification problems, and empirically investigates the relationship between the percentage of neurons left unlabeled and classification accuracy. Practical recommendations are also presented, which address the treatment of unlabeled neurons and the selection of an appropriate neuron labeling algorithm.
{"title":"An investigation into the effect of unlabeled neurons on Self-Organizing Maps","authors":"Willem S. van Heerden, A. Engelbrecht","doi":"10.1109/SSCI.2016.7849938","DOIUrl":"https://doi.org/10.1109/SSCI.2016.7849938","url":null,"abstract":"Self-Organizing Maps (SOMs) are unsupervised neural networks that build data models. Neuron labeling attaches descriptive textual labels to the neurons making up a SOM, and is an important component of SOM-based exploratory data analysis (EDA) and data mining (DM). Several neuron labeling approaches tend to leave some neurons unlabeled. The interaction between unlabeled neurons and SOM model accuracy affect the choice of labeling algorithm for SOM-based EDA and DM, but has not been previously investigated. This paper applies the widely used example-centric neuron labeling algorithm to several classification problems, and empirically investigates the relationship between the percentage of neurons left unlabeled and classification accuracy. Practical recommendations are also presented, which address the treatment of unlabeled neurons and the selection of an appropriate neuron labeling algorithm.","PeriodicalId":120288,"journal":{"name":"2016 IEEE Symposium Series on Computational Intelligence (SSCI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2016-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133455668","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：