Pub Date : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492828
Kee-Hoon Kim, Chang-Seok Lee, Sang-Muk Jo, Sung-Bae Cho
Recently, exploitations of the financial big data to solve the real world problems have been to the fore. Deep neural networks are one of the famous machine learning classifiers as their automatic feature extractions are useful, and even more, their performance is impressive in practical problems. Deep convolutional neural network, one of the promising deep neural networks, can handle the local relationship between their nodes which can make this model powerful in the area of image and speech recognition. In this paper, we propose the deep convolutional neural network architecture that predicts whether a given customer is proper for bank telemarketing or not. The number of layers, learning rate, initial value of nodes, and other parameters that should be set to construct deep convolutional neural network are analyzed and proposed. To validate the proposed model, we use the bank marketing data of 45,211 phone calls collected during 30 months, and attain 76.70% of accuracy which outperforms other conventional classifiers.
{"title":"Predicting the success of bank telemarketing using deep convolutional neural network","authors":"Kee-Hoon Kim, Chang-Seok Lee, Sang-Muk Jo, Sung-Bae Cho","doi":"10.1109/SOCPAR.2015.7492828","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492828","url":null,"abstract":"Recently, exploitations of the financial big data to solve the real world problems have been to the fore. Deep neural networks are one of the famous machine learning classifiers as their automatic feature extractions are useful, and even more, their performance is impressive in practical problems. Deep convolutional neural network, one of the promising deep neural networks, can handle the local relationship between their nodes which can make this model powerful in the area of image and speech recognition. In this paper, we propose the deep convolutional neural network architecture that predicts whether a given customer is proper for bank telemarketing or not. The number of layers, learning rate, initial value of nodes, and other parameters that should be set to construct deep convolutional neural network are analyzed and proposed. To validate the proposed model, we use the bank marketing data of 45,211 phone calls collected during 30 months, and attain 76.70% of accuracy which outperforms other conventional classifiers.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"57 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126537445","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 : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492792
Zhesheng Zhang, Wei Yuan, Fanyu You
This paper considers a system consisting of multiple attackers, lots of charging stations and plug-in electric vehicles (PEVs). The attackers conduct channel jamming attacks and benefit from snatching customers (i.e., PEVs) from the victim charging stations. Suppose that the attackers are selfish and they attempt to maximize their own average net revenue per unit time. We aim to investigate the problem of how to appropriately choose its transmit power to conduct the jamming attack for every attacker. We formulate this problem as a noncooperative game, and show the existence of its solution, i.e., a Nash equilibrium (NE). Due to the existence of some coupling constraints, the game is a Generalized Nash equilibrium problem (GNEP), which is usually hard to solve. To overcome this challenge, here we introduce a variation inequality (VI) approach. More specifically, we treat the game as a VI problem, and prove the existence of its solution. We develop an iterative algorithm to compute the solution to the VI problem, which corresponds to an NE of our game. Numerical results demonstrate the effectiveness and the efficiency of our proposed algorithm.
{"title":"Power allocation of jamming attackers against PEV charging stations: A game theoretical approach","authors":"Zhesheng Zhang, Wei Yuan, Fanyu You","doi":"10.1109/SOCPAR.2015.7492792","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492792","url":null,"abstract":"This paper considers a system consisting of multiple attackers, lots of charging stations and plug-in electric vehicles (PEVs). The attackers conduct channel jamming attacks and benefit from snatching customers (i.e., PEVs) from the victim charging stations. Suppose that the attackers are selfish and they attempt to maximize their own average net revenue per unit time. We aim to investigate the problem of how to appropriately choose its transmit power to conduct the jamming attack for every attacker. We formulate this problem as a noncooperative game, and show the existence of its solution, i.e., a Nash equilibrium (NE). Due to the existence of some coupling constraints, the game is a Generalized Nash equilibrium problem (GNEP), which is usually hard to solve. To overcome this challenge, here we introduce a variation inequality (VI) approach. More specifically, we treat the game as a VI problem, and prove the existence of its solution. We develop an iterative algorithm to compute the solution to the VI problem, which corresponds to an NE of our game. Numerical results demonstrate the effectiveness and the efficiency of our proposed algorithm.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"69 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127747630","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 : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492809
W. Liao, Yi Liang, Yi-Chieh Wu
Text messages in an image usually contain useful information related to the scene, such as location, name, direction or warning. As such, robust and efficient scene text detection has gained increasing attention in the area of computer vision recently. However, most existing scene text detection methods are devised to process Latin-based languages. For the few researches that reported the investigation of Chinese text, the detection rate was inferior to the result for English. In this research, we propose a multilingual scene text detection algorithm for both Chinese and English. The method comprises of four stages: 1. Preprocessing by bilateral filter to make the text region more stable. 2. Extracting candidate text edge and region using Canny edge detector and Maximally Stable Extremal Region (MSER) respectively. Then combine these two features to achieve more robust results. 3. Linking candidate characters: considering both horizontal and vertical direction, character candidates are clustered into text candidates using geometrical constraints. 4. Classifying candidate texts using support vector machine (SVM), to separate text and non-text areas. Experimental results show that the proposed method detects both Chinese and English texts, and achieve satisfactory performance compared to those approaches designed only for English detection.
{"title":"An integrated approach for multilingual scene text detection","authors":"W. Liao, Yi Liang, Yi-Chieh Wu","doi":"10.1109/SOCPAR.2015.7492809","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492809","url":null,"abstract":"Text messages in an image usually contain useful information related to the scene, such as location, name, direction or warning. As such, robust and efficient scene text detection has gained increasing attention in the area of computer vision recently. However, most existing scene text detection methods are devised to process Latin-based languages. For the few researches that reported the investigation of Chinese text, the detection rate was inferior to the result for English. In this research, we propose a multilingual scene text detection algorithm for both Chinese and English. The method comprises of four stages: 1. Preprocessing by bilateral filter to make the text region more stable. 2. Extracting candidate text edge and region using Canny edge detector and Maximally Stable Extremal Region (MSER) respectively. Then combine these two features to achieve more robust results. 3. Linking candidate characters: considering both horizontal and vertical direction, character candidates are clustered into text candidates using geometrical constraints. 4. Classifying candidate texts using support vector machine (SVM), to separate text and non-text areas. Experimental results show that the proposed method detects both Chinese and English texts, and achieve satisfactory performance compared to those approaches designed only for English detection.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"237 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134226101","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 : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492804
Kazuya Nakamura, Hiroshi Kawasaki, S. Ono
Two-dimensional (2D) codes are assumed to be printed on flat planes and subject to distortion when printed on non-rigid materials such as papers and clothes. Although general 2D code decoders correct uniform distortion such as perspective distortion, it is difficult to correct non-uniform and irregular distortion of 2D code itself. To cope with this problem, this paper proposes an agent-based approach to reconstruct 2D code. In this approach, auxiliary lines are given to a 2D code and used to recognize the distortion. First, the proposed method finds 2D code area using feature patterns composed by the auxiliary lines, and looks for finder patterns by Convolutional Neural Network (CNN). Then, many agents simultaneously trace the lines referring various image features and neighborhood agents. Feature weights are optimized by Genetic Algorithm. Experimental results showed that the proposed method has prospects that it can decode distorted 2D code without occlusion.
{"title":"Agent-based two-dimensional barcode decoding robust against non-uniform geometric distortion","authors":"Kazuya Nakamura, Hiroshi Kawasaki, S. Ono","doi":"10.1109/SOCPAR.2015.7492804","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492804","url":null,"abstract":"Two-dimensional (2D) codes are assumed to be printed on flat planes and subject to distortion when printed on non-rigid materials such as papers and clothes. Although general 2D code decoders correct uniform distortion such as perspective distortion, it is difficult to correct non-uniform and irregular distortion of 2D code itself. To cope with this problem, this paper proposes an agent-based approach to reconstruct 2D code. In this approach, auxiliary lines are given to a 2D code and used to recognize the distortion. First, the proposed method finds 2D code area using feature patterns composed by the auxiliary lines, and looks for finder patterns by Convolutional Neural Network (CNN). Then, many agents simultaneously trace the lines referring various image features and neighborhood agents. Feature weights are optimized by Genetic Algorithm. Experimental results showed that the proposed method has prospects that it can decode distorted 2D code without occlusion.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"17 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"122948338","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 : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492816
S. I. Suliman, G. Kendall, I. Musirin
Having an effective method for frequency assignment is very important in mobile communications. Artificial Immune Systems, a relatively new methodology for solving optimization-related problems, is investigated in this study to design a cellular frequency assignment model. In this framework, mobile hosts are allowed to continue interacting with the cells that they are currently in, even when the number of frequencies is insufficient. When the number of users increases, it is almost impossible to foresee future frequency demands in a cell. Often cellular networks are faced with insufficient frequencies in cells, whilst other cells have more frequencies available than they need. Our model works by allocating the unused frequencies from the cells with less demand to the overloaded cells. The model demonstrates that it manages to reuse the available free-conflict frequencies efficiently.
{"title":"An effective AIS-based model for frequency assignment in mobile communication","authors":"S. I. Suliman, G. Kendall, I. Musirin","doi":"10.1109/SOCPAR.2015.7492816","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492816","url":null,"abstract":"Having an effective method for frequency assignment is very important in mobile communications. Artificial Immune Systems, a relatively new methodology for solving optimization-related problems, is investigated in this study to design a cellular frequency assignment model. In this framework, mobile hosts are allowed to continue interacting with the cells that they are currently in, even when the number of frequencies is insufficient. When the number of users increases, it is almost impossible to foresee future frequency demands in a cell. Often cellular networks are faced with insufficient frequencies in cells, whilst other cells have more frequencies available than they need. Our model works by allocating the unused frequencies from the cells with less demand to the overloaded cells. The model demonstrates that it manages to reuse the available free-conflict frequencies efficiently.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"3 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114081419","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 : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492788
T. Ahmad, Doni S. Pambudi, T. Usagawa
Biometrics, especially fingerprint, has been popular to use for authenticating users because it is relatively permanence. This characteristic, however, is a problem because once it is compromised, fingerprint data can not be replaced. The conventional cryptographic algorithm may not be able to protect fingerprint data since the fingerprint scanning result is unstable. This paper improves the performance of the previous projection-based fingerprint protection method by removing the need of the core point and applying further minutiae checking hierarchically. The experimental result which is done in a public database produces the EER of about 1%.
{"title":"Improving the performance of projection-based cancelable fingerprint template method","authors":"T. Ahmad, Doni S. Pambudi, T. Usagawa","doi":"10.1109/SOCPAR.2015.7492788","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492788","url":null,"abstract":"Biometrics, especially fingerprint, has been popular to use for authenticating users because it is relatively permanence. This characteristic, however, is a problem because once it is compromised, fingerprint data can not be replaced. The conventional cryptographic algorithm may not be able to protect fingerprint data since the fingerprint scanning result is unstable. This paper improves the performance of the previous projection-based fingerprint protection method by removing the need of the core point and applying further minutiae checking hierarchically. The experimental result which is done in a public database produces the EER of about 1%.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"18 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"125232646","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 : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492772
Rodrigo Lisbôa Pereira, Edson Koiti Kudo Yasojima, R. M. Oliveira, M. A. F. Mollinetti, O. N. Teixeira, R. C. L. Oliveira
The following paper introduces a parallel approach to a social variant of the Genetic Algorithm, called Parallel Genetic Algorithm with Social Interaction (PSIGA). The algorithm is based on social games involving game theory, and it is implemented using the OpenMP API, which is based on the shared memory programming model for multiple processor architectures. The main contribution of this approach is the parallelization using the Shared Memory of the Social Interaction Genetic Algorithm (SIGA) in order to achieve faster and better optimality than its nonparallel counterpart for global optimization problems with restrictions. For means of performance assessment, the algorithm is tested on four instances of engineering design problems and the obtained results compared with the Genetic Algorithm with Social Interaction (SIGA) implemented in sequential programming model.
{"title":"Parallel genetic algorithm with social interaction for solving constrained global optimization problems","authors":"Rodrigo Lisbôa Pereira, Edson Koiti Kudo Yasojima, R. M. Oliveira, M. A. F. Mollinetti, O. N. Teixeira, R. C. L. Oliveira","doi":"10.1109/SOCPAR.2015.7492772","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492772","url":null,"abstract":"The following paper introduces a parallel approach to a social variant of the Genetic Algorithm, called Parallel Genetic Algorithm with Social Interaction (PSIGA). The algorithm is based on social games involving game theory, and it is implemented using the OpenMP API, which is based on the shared memory programming model for multiple processor architectures. The main contribution of this approach is the parallelization using the Shared Memory of the Social Interaction Genetic Algorithm (SIGA) in order to achieve faster and better optimality than its nonparallel counterpart for global optimization problems with restrictions. For means of performance assessment, the algorithm is tested on four instances of engineering design problems and the obtained results compared with the Genetic Algorithm with Social Interaction (SIGA) implemented in sequential programming model.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"127818739","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 : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492781
Esraa Elhariri, Nashwa El-Bendary, A. Hassanien, A. Abraham
Grey Wolf Optimization (GWO) algorithm is a new meta-heuristic method, which is inspired by grey wolves, to mimic the hierarchy of leadership and grey wolves hunting mechanism in nature. This paper presents a hybrid model that employs grey wolf optimizer (GWO) along with support vector machines (SVMs) classification algorithm to improve the classification accuracy via selecting the optimal settings of SVMs parameters. The proposed approach consists of three phases; namely pre-processing, feature extraction, and GWO-SVMs classification phases. The proposed classification approach was implemented by applying resizing, remove background, and extracting color components for each image. Then, feature vector generation has been implemented via applying PCA feature extraction. Finally, GWO-SVMs model is developed for selecting the optimal SVMs parameters. The proposed approach has been implemented via applying One-againstOne multi-class SVMs system using 3-fold cross-validation. The datasets used for experiments were constructed based on real sample images of bell pepper at different stages, which were collected from farms in Minya city, Upper Egypt. Datasets of total 175 images were used for both training and testing datasets. Experimental results indicated that the proposed GWO-SVMs approach achieved better classification accuracy compared to the typical SVMs classification algorithm.
{"title":"Grey wolf optimization for one-against-one multi-class support vector machines","authors":"Esraa Elhariri, Nashwa El-Bendary, A. Hassanien, A. Abraham","doi":"10.1109/SOCPAR.2015.7492781","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492781","url":null,"abstract":"Grey Wolf Optimization (GWO) algorithm is a new meta-heuristic method, which is inspired by grey wolves, to mimic the hierarchy of leadership and grey wolves hunting mechanism in nature. This paper presents a hybrid model that employs grey wolf optimizer (GWO) along with support vector machines (SVMs) classification algorithm to improve the classification accuracy via selecting the optimal settings of SVMs parameters. The proposed approach consists of three phases; namely pre-processing, feature extraction, and GWO-SVMs classification phases. The proposed classification approach was implemented by applying resizing, remove background, and extracting color components for each image. Then, feature vector generation has been implemented via applying PCA feature extraction. Finally, GWO-SVMs model is developed for selecting the optimal SVMs parameters. The proposed approach has been implemented via applying One-againstOne multi-class SVMs system using 3-fold cross-validation. The datasets used for experiments were constructed based on real sample images of bell pepper at different stages, which were collected from farms in Minya city, Upper Egypt. Datasets of total 175 images were used for both training and testing datasets. Experimental results indicated that the proposed GWO-SVMs approach achieved better classification accuracy compared to the typical SVMs classification algorithm.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"42 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130692553","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 : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492770
Wai Chong Chia, L. Yeong, S. I. Ch'ng, Yoke Lun Kam
In this paper, the effect of using multi-image and single-image super-resolution to reduce registration errors of low resolution images is evaluated. Two sets of low resolution images were captured using CMUCam4 to perform the evaluation. Moreover, a simplified method that make use of feature points extracted from resolution enhanced / upscaled images to improve the registration of low resolution images is also presented. The simulation results show that enhancing / upscaling the images in prior to registration does help to reduce the registration errors.
{"title":"The effect of using super-resolution to improve feature extraction and registration of low resolution images in sensor networks","authors":"Wai Chong Chia, L. Yeong, S. I. Ch'ng, Yoke Lun Kam","doi":"10.1109/SOCPAR.2015.7492770","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492770","url":null,"abstract":"In this paper, the effect of using multi-image and single-image super-resolution to reduce registration errors of low resolution images is evaluated. Two sets of low resolution images were captured using CMUCam4 to perform the evaluation. Moreover, a simplified method that make use of feature points extracted from resolution enhanced / upscaled images to improve the registration of low resolution images is also presented. The simulation results show that enhancing / upscaling the images in prior to registration does help to reduce the registration errors.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"31 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130980786","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 : 2015-11-01DOI: 10.1109/SOCPAR.2015.7492774
Takaya Ogiso, K. Yamauchi, Norio Ishii, Yuri Suzuki
Artificial intelligence systems are frequently used to solve various problems in our daily lives. However, these systems require problem-specific big data to facilitate their learning processes. Unfortunately, for unknown environments, there are no previous instances available for learning. To support such learning in unknown environments, we propose a novel hybrid learning system that facilitates collaborative learning between humans and artificial intelligence systems. In this study, we verified that the proposed system accelerated the both human and machine learning by employing a simplified color design task.
{"title":"A co-learning system for humans and machines","authors":"Takaya Ogiso, K. Yamauchi, Norio Ishii, Yuri Suzuki","doi":"10.1109/SOCPAR.2015.7492774","DOIUrl":"https://doi.org/10.1109/SOCPAR.2015.7492774","url":null,"abstract":"Artificial intelligence systems are frequently used to solve various problems in our daily lives. However, these systems require problem-specific big data to facilitate their learning processes. Unfortunately, for unknown environments, there are no previous instances available for learning. To support such learning in unknown environments, we propose a novel hybrid learning system that facilitates collaborative learning between humans and artificial intelligence systems. In this study, we verified that the proposed system accelerated the both human and machine learning by employing a simplified color design task.","PeriodicalId":409493,"journal":{"name":"2015 7th International Conference of Soft Computing and Pattern Recognition (SoCPaR)","volume":"5 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2015-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"133038402","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：