This paper describes methods for controlling routing points of VLAN domains using binary particle swarm optimization (BPSO) and angle modulated particle swarm optimization (AMPSO). Virtual LAN (VLAN) is a technique for virtualizing data link layer (or L2) and can construct arbitrary logical networks on top of a physical network. However, VLAN often causes much redundant traffic due to inappropriate deployments of network-layer (L3) routing capabilities in VLAN networks. We propose two methods using BPSO and AMPSO, and show that they can adaptively select the routing points dynamically in accordance with the observed traffic patterns and thus reduce the redundant traffic. The convergence features are compared with those of the conventional method on the basis of a statistical method. Then we also show that the scalability of the algorithm using AMPOS is high and thus we can expect that it is applicable to practical large VLAN environments.
{"title":"Adaptive Routing Point Control in Virtualized Local Area Networks Using Particle Swarm Optimizations","authors":"Kensuke Takahashi, Toshio Hirotsu, T. Sugawara","doi":"10.1109/ICTAI.2011.59","DOIUrl":"https://doi.org/10.1109/ICTAI.2011.59","url":null,"abstract":"This paper describes methods for controlling routing points of VLAN domains using binary particle swarm optimization (BPSO) and angle modulated particle swarm optimization (AMPSO). Virtual LAN (VLAN) is a technique for virtualizing data link layer (or L2) and can construct arbitrary logical networks on top of a physical network. However, VLAN often causes much redundant traffic due to inappropriate deployments of network-layer (L3) routing capabilities in VLAN networks. We propose two methods using BPSO and AMPSO, and show that they can adaptively select the routing points dynamically in accordance with the observed traffic patterns and thus reduce the redundant traffic. The convergence features are compared with those of the conventional method on the basis of a statistical method. Then we also show that the scalability of the algorithm using AMPOS is high and thus we can expect that it is applicable to practical large VLAN environments.","PeriodicalId":332661,"journal":{"name":"2011 IEEE 23rd International Conference on Tools with Artificial Intelligence","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115815909","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}
Stefano Bistarelli, Giorgio Gosti, Francesco Santini
In this paper we focus on solving Fuzzy Distributes Constraint Satisfaction Problems (Fuzzy DCSPs) with an algorithm for Naming Games (NGs): each word on which the agents have to agree on is associated with a preference represented as a fuzzy score. The solution is the agreed word associated with the highest preference value. The two main features that distinguish this methodology from Fuzzy DCSPs methods are that the system can react to small instance changes and and it does not require pre-agreed agent/variable ordering.
{"title":"Solving Fuzzy DCSPs with Naming Games","authors":"Stefano Bistarelli, Giorgio Gosti, Francesco Santini","doi":"10.1109/ICTAI.2011.159","DOIUrl":"https://doi.org/10.1109/ICTAI.2011.159","url":null,"abstract":"In this paper we focus on solving Fuzzy Distributes Constraint Satisfaction Problems (Fuzzy DCSPs) with an algorithm for Naming Games (NGs): each word on which the agents have to agree on is associated with a preference represented as a fuzzy score. The solution is the agreed word associated with the highest preference value. The two main features that distinguish this methodology from Fuzzy DCSPs methods are that the system can react to small instance changes and and it does not require pre-agreed agent/variable ordering.","PeriodicalId":332661,"journal":{"name":"2011 IEEE 23rd International Conference on Tools with Artificial Intelligence","volume":"33 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125169396","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}
RELIEF is a very effective and extremely popular feature selection algorithm developed for the first time in 1992 by Kira and Rendell. Since then it has been modified and expanded in various ways to make it more efficient. But the original RELIEF and all of its expansions are for feature selection over labeled data for classification purposes. To the best of our knowledge, for the first time ever RELIEF is used in this paper as RELIEF-C for unlabeled data to select relevant features for clustering. We modified RELIEF so as to overcome its inherent difficulties in the presence of large number of irrelevant features and/or significant number of noisy tuples. RELIEF-C has several advantages over existing wrapper and filter feature selection methods: (a) it works well in the presence of large amount of noisy tuples, (b) it is robust even when underlying clustering algorithm fails to cluster properly, and (c) it accurately recognizes the relevant features even in the presence of large number of irrelevant features. We compared RELIEF-C with two established feature selection methods for clustering. RELIEF-C outperforms other methods significantly over synthetic, benchmark and real world data sets particularly when data set consists of large amount of noisy tuples and/or irrelevant features.
{"title":"RELIEF-C: Efficient Feature Selection for Clustering over Noisy Data","authors":"M. Dash, Y. Ong","doi":"10.1109/ICTAI.2011.135","DOIUrl":"https://doi.org/10.1109/ICTAI.2011.135","url":null,"abstract":"RELIEF is a very effective and extremely popular feature selection algorithm developed for the first time in 1992 by Kira and Rendell. Since then it has been modified and expanded in various ways to make it more efficient. But the original RELIEF and all of its expansions are for feature selection over labeled data for classification purposes. To the best of our knowledge, for the first time ever RELIEF is used in this paper as RELIEF-C for unlabeled data to select relevant features for clustering. We modified RELIEF so as to overcome its inherent difficulties in the presence of large number of irrelevant features and/or significant number of noisy tuples. RELIEF-C has several advantages over existing wrapper and filter feature selection methods: (a) it works well in the presence of large amount of noisy tuples, (b) it is robust even when underlying clustering algorithm fails to cluster properly, and (c) it accurately recognizes the relevant features even in the presence of large number of irrelevant features. We compared RELIEF-C with two established feature selection methods for clustering. RELIEF-C outperforms other methods significantly over synthetic, benchmark and real world data sets particularly when data set consists of large amount of noisy tuples and/or irrelevant features.","PeriodicalId":332661,"journal":{"name":"2011 IEEE 23rd International Conference on Tools with Artificial Intelligence","volume":"39 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114164839","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}
S. Schneider, A. Melkumyan, R. Murphy, E. Nettleton
In this paper we use a machine learning algorithm based on Gaussian Processes (GPs) and the Observation Angle Dependent (OAD) covariance function to classify hyper spectral imagery for the first time. This paper demonstrates the potential of the GP-OAD method for use in autonomous mining to identify and map geology and mineralogy on a vertical mine face. We discuss the importance of independent training data (i.e. a spectral library) to map any mine face without a priori knowledge. We compare an independent spectral library to other libraries, based on image data, and evaluate their relative performances to distinguish ore bearing zones from waste. Results show that the algorithm yields high accuracies (90%) and F-scores (77%), the best results are achieved when libraries are combined. We also demonstrate mapping of geology using imagery under different conditions of illumination (e.g. shade).
{"title":"Classification of Hyperspectral Imagery Using GPs and the OAD Covariance Function with Automated Endmember Extraction","authors":"S. Schneider, A. Melkumyan, R. Murphy, E. Nettleton","doi":"10.1109/ICTAI.2011.189","DOIUrl":"https://doi.org/10.1109/ICTAI.2011.189","url":null,"abstract":"In this paper we use a machine learning algorithm based on Gaussian Processes (GPs) and the Observation Angle Dependent (OAD) covariance function to classify hyper spectral imagery for the first time. This paper demonstrates the potential of the GP-OAD method for use in autonomous mining to identify and map geology and mineralogy on a vertical mine face. We discuss the importance of independent training data (i.e. a spectral library) to map any mine face without a priori knowledge. We compare an independent spectral library to other libraries, based on image data, and evaluate their relative performances to distinguish ore bearing zones from waste. Results show that the algorithm yields high accuracies (90%) and F-scores (77%), the best results are achieved when libraries are combined. We also demonstrate mapping of geology using imagery under different conditions of illumination (e.g. shade).","PeriodicalId":332661,"journal":{"name":"2011 IEEE 23rd International Conference on Tools with Artificial Intelligence","volume":"105 8","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121014123","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}
Trajectory data streams are huge amounts of data pertaining to time and position of moving objects. They are continuously generated by different sources exploiting a wide variety of technologies (e.g., RFID tags, GPS, GSM networks). Mining such amounts of data is challenging, since the possibility to extract useful information from this peculiar kind of data is crucial in many application scenarios such as vehicle traffic management, hand-off in cellular networks, supply chain management. Moreover, spatial data poses interesting challenges both for their proper definition and acquisition, thus making the mining process harder than for classical point data. In this paper, we address the problem of trajectory data outlier detection, that revealed really challenging as we deal with data (trajectories) for which the order of elements is relevant. We propose a complete framework starting from data preparation task that allows us to make the mining step quite effective. Since the validation of data mining approaches has to be experimental we performed several tests on real world datasets that confirmed the efficiency and effectiveness of the proposed technique.
{"title":"Trajectory Outlier Detection Using an Analytical Approach","authors":"E. Masciari","doi":"10.1109/ICTAI.2011.62","DOIUrl":"https://doi.org/10.1109/ICTAI.2011.62","url":null,"abstract":"Trajectory data streams are huge amounts of data pertaining to time and position of moving objects. They are continuously generated by different sources exploiting a wide variety of technologies (e.g., RFID tags, GPS, GSM networks). Mining such amounts of data is challenging, since the possibility to extract useful information from this peculiar kind of data is crucial in many application scenarios such as vehicle traffic management, hand-off in cellular networks, supply chain management. Moreover, spatial data poses interesting challenges both for their proper definition and acquisition, thus making the mining process harder than for classical point data. In this paper, we address the problem of trajectory data outlier detection, that revealed really challenging as we deal with data (trajectories) for which the order of elements is relevant. We propose a complete framework starting from data preparation task that allows us to make the mining step quite effective. Since the validation of data mining approaches has to be experimental we performed several tests on real world datasets that confirmed the efficiency and effectiveness of the proposed technique.","PeriodicalId":332661,"journal":{"name":"2011 IEEE 23rd International Conference on Tools with Artificial Intelligence","volume":"329 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115975146","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}
A swarm of robots deployed in dynamic, hostile environments may encounter situations that can prevent them from achieving optimality or completing certain tasks. To resolve these situations, the robots must have an adaptive software system that can proactively cope with changes. This adaptive system should emulate the intelligence of human reasoning and common sense but must not assume that the robots can communicate, be tightly coupled, or be constantly at a close range. This paper presents a path strategy evaluator (PSE) that learns an optimal path by considering not just the distance, but also how to minimize damages to each robot and enhance the likelihood that the swarm will succeed in its mission, all with minimal impositions on the functionality of the robots. Our evaluation shows that this PSE is able to learn a dynamic environment and its effect on the robots' critical components and output an optimal path for the robots.
{"title":"ROBUST Path Strategy Evaluator","authors":"Angie Shia, F. Bastani, I. Yen","doi":"10.1109/ICTAI.2011.91","DOIUrl":"https://doi.org/10.1109/ICTAI.2011.91","url":null,"abstract":"A swarm of robots deployed in dynamic, hostile environments may encounter situations that can prevent them from achieving optimality or completing certain tasks. To resolve these situations, the robots must have an adaptive software system that can proactively cope with changes. This adaptive system should emulate the intelligence of human reasoning and common sense but must not assume that the robots can communicate, be tightly coupled, or be constantly at a close range. This paper presents a path strategy evaluator (PSE) that learns an optimal path by considering not just the distance, but also how to minimize damages to each robot and enhance the likelihood that the swarm will succeed in its mission, all with minimal impositions on the functionality of the robots. Our evaluation shows that this PSE is able to learn a dynamic environment and its effect on the robots' critical components and output an optimal path for the robots.","PeriodicalId":332661,"journal":{"name":"2011 IEEE 23rd International Conference on Tools with Artificial Intelligence","volume":"11 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115224225","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}
Ability to monitor and detect abnormalities accurately is important in a manufacturing process. This can be achieved by recognizing abnormalities in its control charts. This work is concerned with classification of control chart patterns (CCPs) by utilizing a technique known as Symbolic Aggregate Approximation (SAX) and an evolutionary based data mining program known as Self-adjusting Association Rules Generator (SARG). SAX is used in preprocessing to transform CCPs, which can be considered as time series, to symbolic representations. SARG is then applied to these symbolic representations to generate a classifier in a form of a nested IF-THEN-ELSE rules. A more efficient nested IF-THEN-ELSE rules classifier in SARG is discovered. A systematic investigation was carried out to find the capability of the proposed method. This was done by attempting to generate classifiers for CCPs datasets with different level of noises in them. CCPs were generated by Generalized Autoregressive Conditional Heteroskedasticity (GARH) Model where ó is the noise level parameter. Two crucial parameters in SAX are Piecewise Aggregate Approximation and Alphabet Size values. This work identifies suitable values for both parameters in SAX for SARG to generate CCPs classifiers. This is the first work to generate CCPs classifiers with accuracy up to 90% for ó at 13 and 95 % for ó at 9.
{"title":"Capability of Classification of Control Chart Patterns Classifiers Using Symbolic Representation Preprocessing and Evolutionary Computation","authors":"K. Lavangnananda, P. Sawasdimongkol","doi":"10.1109/ICTAI.2011.178","DOIUrl":"https://doi.org/10.1109/ICTAI.2011.178","url":null,"abstract":"Ability to monitor and detect abnormalities accurately is important in a manufacturing process. This can be achieved by recognizing abnormalities in its control charts. This work is concerned with classification of control chart patterns (CCPs) by utilizing a technique known as Symbolic Aggregate Approximation (SAX) and an evolutionary based data mining program known as Self-adjusting Association Rules Generator (SARG). SAX is used in preprocessing to transform CCPs, which can be considered as time series, to symbolic representations. SARG is then applied to these symbolic representations to generate a classifier in a form of a nested IF-THEN-ELSE rules. A more efficient nested IF-THEN-ELSE rules classifier in SARG is discovered. A systematic investigation was carried out to find the capability of the proposed method. This was done by attempting to generate classifiers for CCPs datasets with different level of noises in them. CCPs were generated by Generalized Autoregressive Conditional Heteroskedasticity (GARH) Model where ó is the noise level parameter. Two crucial parameters in SAX are Piecewise Aggregate Approximation and Alphabet Size values. This work identifies suitable values for both parameters in SAX for SARG to generate CCPs classifiers. This is the first work to generate CCPs classifiers with accuracy up to 90% for ó at 13 and 95 % for ó at 9.","PeriodicalId":332661,"journal":{"name":"2011 IEEE 23rd International Conference on Tools with Artificial Intelligence","volume":"78 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124736107","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}
Reading is an integral part of educational development, however, it is frustrating for people who struggle to understand (are not motivated to read, respectively) text documents that are beyond (below, respectively) their readability levels. Finding appropriate reading materials, with or without first scanning through their contents, is a challenge, since there are tremendous amount of documents these days and a clear majority of them are not tagged with their readability levels. Even though existing readability assessment tools determine readability levels of text documents, they analyze solely the lexical, syntactic, and/or semantic properties of a document, which are neither fully-automated, generalized, nor well-defined and are mostly based on observations. To advance the current readability analysis technique, we propose a robust, fully-automated readability analyzer, denoted ReadAid, which employs support vector machines to combine features from the US Curriculum and College Board, traditional readability measures, and the author(s) and subject area(s) of a text document d to assess the readability level of d. ReadAid can be applied for (i) filtering documents (retrieved in response to a web query) of a particular readability level, (ii) determining the readability levels of digitalized text documents, such as book chapters, magazine articles, and news stories, or (iii) dynamically analyzing, in real time, the grade level of a text document being created. The novelty of ReadAid lies on using authorship, subject areas, and academic concepts and grammatical constructions extracted from the US Curriculum to determine the readability level of a text document. Experimental results show that ReadAid is highly effective and outperforms existing state-of-the-art readability assessment tools.
{"title":"ReadAid: A Robust and Fully-Automated Readability Assessment Tool","authors":"Rani Qumsiyeh, Yiu-Kai Ng","doi":"10.1109/ICTAI.2011.87","DOIUrl":"https://doi.org/10.1109/ICTAI.2011.87","url":null,"abstract":"Reading is an integral part of educational development, however, it is frustrating for people who struggle to understand (are not motivated to read, respectively) text documents that are beyond (below, respectively) their readability levels. Finding appropriate reading materials, with or without first scanning through their contents, is a challenge, since there are tremendous amount of documents these days and a clear majority of them are not tagged with their readability levels. Even though existing readability assessment tools determine readability levels of text documents, they analyze solely the lexical, syntactic, and/or semantic properties of a document, which are neither fully-automated, generalized, nor well-defined and are mostly based on observations. To advance the current readability analysis technique, we propose a robust, fully-automated readability analyzer, denoted ReadAid, which employs support vector machines to combine features from the US Curriculum and College Board, traditional readability measures, and the author(s) and subject area(s) of a text document d to assess the readability level of d. ReadAid can be applied for (i) filtering documents (retrieved in response to a web query) of a particular readability level, (ii) determining the readability levels of digitalized text documents, such as book chapters, magazine articles, and news stories, or (iii) dynamically analyzing, in real time, the grade level of a text document being created. The novelty of ReadAid lies on using authorship, subject areas, and academic concepts and grammatical constructions extracted from the US Curriculum to determine the readability level of a text document. Experimental results show that ReadAid is highly effective and outperforms existing state-of-the-art readability assessment tools.","PeriodicalId":332661,"journal":{"name":"2011 IEEE 23rd International Conference on Tools with Artificial Intelligence","volume":"41 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128351711","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}
Personalized systems are a response to the increasing number of resources on the Internet. In order to facilitate their design and creation, we aim at formalizing them. In this paper, we consider the relationship between a personalized application and its non-personalized counterpart. We argue that a personalized application is a formal extension of a non-personalized one. We aim at characterizing the syntactic differences between the expression of the personalized and non-personalized versions of the application. Situation calculus is our framework to formalize applications. We introduce two scenarios of non-personalized application that we personalize to illustrate our approach.
{"title":"A Formal Approach to Personalization","authors":"G. Dubus, Fabrice Popineau, Yolaine Bourda","doi":"10.1109/ICTAI.2011.43","DOIUrl":"https://doi.org/10.1109/ICTAI.2011.43","url":null,"abstract":"Personalized systems are a response to the increasing number of resources on the Internet. In order to facilitate their design and creation, we aim at formalizing them. In this paper, we consider the relationship between a personalized application and its non-personalized counterpart. We argue that a personalized application is a formal extension of a non-personalized one. We aim at characterizing the syntactic differences between the expression of the personalized and non-personalized versions of the application. Situation calculus is our framework to formalize applications. We introduce two scenarios of non-personalized application that we personalize to illustrate our approach.","PeriodicalId":332661,"journal":{"name":"2011 IEEE 23rd International Conference on Tools with Artificial Intelligence","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128841814","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}
Jicheng Fu, Jerrad Genson, Yih-Kuen Jan, Maria Jones
People with spinal cord injury (SCI) are at risk for pressure ulcers because of their poor motor function and consequent prolonged sitting in wheelchairs. The current clinical practice typically uses the wheelchair tilt and recline to attain specific seating angles (sitting postures) to reduce seating pressure in order to prevent pressure ulcers. The rationale is to allow the development of reactive hyperemia to re-perfuse the ischemic tissues. However, our study reveals that a particular tilt and recline setting may result in a significant increase of skin perfusion for one person with SCI, but may cause neutral or even negative effect on another person. Therefore, an individualized guidance on wheelchair tilt and recline usage is desirable in people with various levels of SCI. In this study, we intend to demonstrate the feasibility of using machine-learning techniques to classify and predict favorable wheelchair tilt and recline settings for individual wheelchair users with SCI. Specifically, we use artificial neural networks (ANNs) to classify whether a given tilt and recline setting would cause a positive, neutral, or negative skin perfusion response. The challenge, however, is that ANN is prone to over fitting, a situation in which ANN can perfectly classify the existing data while cannot correctly classify new (unseen) data. We investigate using the genetic algorithm (GA) to train ANN to reduce the chance of converging on local optima and improve the generalization capability of classifying unseen data. Our experimental results indicate that the GA-based ANN significantly improves the generalization ability and outperforms the traditional statistical approach and other commonly used classification techniques, such as BP-based ANN and support vector machine (SVM). To the best of our knowledge, there are no such intelligent systems available now. Our research fills in the gap in existing evidence.
{"title":"Using Artificial Neural Network to Determine Favorable Wheelchair Tilt and Recline Usage in People with Spinal Cord Injury: Training ANN with Genetic Algorithm to Improve Generalization","authors":"Jicheng Fu, Jerrad Genson, Yih-Kuen Jan, Maria Jones","doi":"10.1109/ICTAI.2011.13","DOIUrl":"https://doi.org/10.1109/ICTAI.2011.13","url":null,"abstract":"People with spinal cord injury (SCI) are at risk for pressure ulcers because of their poor motor function and consequent prolonged sitting in wheelchairs. The current clinical practice typically uses the wheelchair tilt and recline to attain specific seating angles (sitting postures) to reduce seating pressure in order to prevent pressure ulcers. The rationale is to allow the development of reactive hyperemia to re-perfuse the ischemic tissues. However, our study reveals that a particular tilt and recline setting may result in a significant increase of skin perfusion for one person with SCI, but may cause neutral or even negative effect on another person. Therefore, an individualized guidance on wheelchair tilt and recline usage is desirable in people with various levels of SCI. In this study, we intend to demonstrate the feasibility of using machine-learning techniques to classify and predict favorable wheelchair tilt and recline settings for individual wheelchair users with SCI. Specifically, we use artificial neural networks (ANNs) to classify whether a given tilt and recline setting would cause a positive, neutral, or negative skin perfusion response. The challenge, however, is that ANN is prone to over fitting, a situation in which ANN can perfectly classify the existing data while cannot correctly classify new (unseen) data. We investigate using the genetic algorithm (GA) to train ANN to reduce the chance of converging on local optima and improve the generalization capability of classifying unseen data. Our experimental results indicate that the GA-based ANN significantly improves the generalization ability and outperforms the traditional statistical approach and other commonly used classification techniques, such as BP-based ANN and support vector machine (SVM). To the best of our knowledge, there are no such intelligent systems available now. Our research fills in the gap in existing evidence.","PeriodicalId":332661,"journal":{"name":"2011 IEEE 23rd International Conference on Tools with Artificial Intelligence","volume":"6 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2011-11-07","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129637572","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：