Pub Date : 2018-11-01DOI: 10.1109/ICSAI.2018.8599343
Z. Chang, Hongmei Sun, Wenyuan Zhao, Yongjing Liu, Ning Zhao, Xiaole Han, Qiang Zhang, Zichuang Yan, Cheng Wu, Yunzhen Wei
lung adenocarcinoma (LUAD) is the highest incidence of malignant tumors in China, because its early symptoms are not easily ignored by most people, so screening reliable markers is the focus of research. In order to screen stable, cross-platform prognostic lncRNA of LUAD, we combined gene expression profiles and RNAseq data to screen stable lncRNA pairs shared by both platforms and significantly reversed lncRNA pairs in LUAD samples, and then screened the prognostic-related lncRNA from the reversed lncRNA pairs. The stability of the gene pairs obtained from the normal samples was 98.72%. Twenty pairs of survival-related significantly reversed lncRNA gene pairs with 33 lncRNA were found. The results of functional enrichment showed that the functions of these lncRNA pairs were consistent with the known results. Because of its cross-platform and easy detection of independent samples, this result can be better applied to clinical diagnosis and can detect the risk and prognosis of possible LUAD in advance.
{"title":"Identification of Prognostic Markers for LUAD based on Rank Expression of Long non-coding RNA","authors":"Z. Chang, Hongmei Sun, Wenyuan Zhao, Yongjing Liu, Ning Zhao, Xiaole Han, Qiang Zhang, Zichuang Yan, Cheng Wu, Yunzhen Wei","doi":"10.1109/ICSAI.2018.8599343","DOIUrl":"https://doi.org/10.1109/ICSAI.2018.8599343","url":null,"abstract":"lung adenocarcinoma (LUAD) is the highest incidence of malignant tumors in China, because its early symptoms are not easily ignored by most people, so screening reliable markers is the focus of research. In order to screen stable, cross-platform prognostic lncRNA of LUAD, we combined gene expression profiles and RNAseq data to screen stable lncRNA pairs shared by both platforms and significantly reversed lncRNA pairs in LUAD samples, and then screened the prognostic-related lncRNA from the reversed lncRNA pairs. The stability of the gene pairs obtained from the normal samples was 98.72%. Twenty pairs of survival-related significantly reversed lncRNA gene pairs with 33 lncRNA were found. The results of functional enrichment showed that the functions of these lncRNA pairs were consistent with the known results. Because of its cross-platform and easy detection of independent samples, this result can be better applied to clinical diagnosis and can detect the risk and prognosis of possible LUAD in advance.","PeriodicalId":375852,"journal":{"name":"2018 5th International Conference on Systems and Informatics (ICSAI)","volume":"109 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126898337","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}
The traditional signature-based malware detection technology, which restricted by the updating frequency of the feature dataset, that cannot identify the new malware sample quickly. Malware from same type or same family usually have similar behaviors. Therefore, by comparing the similarity between the sequences represented by the function call sequence, which is less affected by the update frequency of the feature dataset. However, in face of a large number of malicious code samples to be detected, the size of the sequences extracted from the samples increases exponentially, which cannot guarantee the real-time detection of malware. In order to ensure the real time of malicious code detection, a parallel method based malicious code sequence comparison model is proposed in this paper. It includes two levels of parallelism, representing parallelism of different granularity, which effectively improves the efficiency of malicious code detection and recognition. The evaluation shows that our method has high effectiveness and efficiency with the large-scale data sets.
{"title":"Malware Detection and Classification Based on Parallel Sequence Comparison","authors":"Hao Ding, Wenjie Sun, Yihang Chen, Bing-lin Zhao, Hairen Gui","doi":"10.1109/ICSAI.2018.8599509","DOIUrl":"https://doi.org/10.1109/ICSAI.2018.8599509","url":null,"abstract":"The traditional signature-based malware detection technology, which restricted by the updating frequency of the feature dataset, that cannot identify the new malware sample quickly. Malware from same type or same family usually have similar behaviors. Therefore, by comparing the similarity between the sequences represented by the function call sequence, which is less affected by the update frequency of the feature dataset. However, in face of a large number of malicious code samples to be detected, the size of the sequences extracted from the samples increases exponentially, which cannot guarantee the real-time detection of malware. In order to ensure the real time of malicious code detection, a parallel method based malicious code sequence comparison model is proposed in this paper. It includes two levels of parallelism, representing parallelism of different granularity, which effectively improves the efficiency of malicious code detection and recognition. The evaluation shows that our method has high effectiveness and efficiency with the large-scale data sets.","PeriodicalId":375852,"journal":{"name":"2018 5th International Conference on Systems and Informatics (ICSAI)","volume":"59 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121107195","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 : 2018-11-01DOI: 10.1109/ICSAI.2018.8599461
Haochen Shi, Haipeng Xiao, Jianjiang Zhou, Ning Li, Huiyu Zhou
By analyzing the influences of kernel parameter and penalty factor for generalization performance on Support Vector Machine (SVM), a novel parameter optimization algorithm based on segmented dichotomy is proposed for Radial Basis Function (RBF) kernel. Combine with Segmented Dichotomy(SD) and Gird Searching(GS) method, a composite parameter selection, SD-GS algorithm, is structured for rapid optimization of kernel parameter and penalty factor. UCI Machine Learning database is used to test our proposed method. Experimental results have shown that performance on parameter selection is better than traversal exponential grid searching. Thus, the optimized parameter combination of SD-GS algorithm enables RBF kernel in SVM to have higher generalization performance.
{"title":"Radial Basis Function Kernel Parameter Optimization Algorithm in Support Vector Machine Based on Segmented Dichotomy","authors":"Haochen Shi, Haipeng Xiao, Jianjiang Zhou, Ning Li, Huiyu Zhou","doi":"10.1109/ICSAI.2018.8599461","DOIUrl":"https://doi.org/10.1109/ICSAI.2018.8599461","url":null,"abstract":"By analyzing the influences of kernel parameter and penalty factor for generalization performance on Support Vector Machine (SVM), a novel parameter optimization algorithm based on segmented dichotomy is proposed for Radial Basis Function (RBF) kernel. Combine with Segmented Dichotomy(SD) and Gird Searching(GS) method, a composite parameter selection, SD-GS algorithm, is structured for rapid optimization of kernel parameter and penalty factor. UCI Machine Learning database is used to test our proposed method. Experimental results have shown that performance on parameter selection is better than traversal exponential grid searching. Thus, the optimized parameter combination of SD-GS algorithm enables RBF kernel in SVM to have higher generalization performance.","PeriodicalId":375852,"journal":{"name":"2018 5th International Conference on Systems and Informatics (ICSAI)","volume":"190 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114855209","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 : 2018-11-01DOI: 10.1109/ICSAI.2018.8599475
Krystian Wojtkiewicz
A new approach to modeling of various processes, namely Unified Process Metamodel, is presented. This universal solution serves the needs of modeling of modern decision-making systems used in a number of scientific areas. The metamodel uses solutions based on various knowledge engineering techniques. Four functional layers are selected under a new method, in each of them the specific types of components are defined. The particular attention to resource flow and control layer is devoted in the paper.
{"title":"Event Condition Action Approach to Process’ Control Layer Modeling in Unified Process Metamodel","authors":"Krystian Wojtkiewicz","doi":"10.1109/ICSAI.2018.8599475","DOIUrl":"https://doi.org/10.1109/ICSAI.2018.8599475","url":null,"abstract":"A new approach to modeling of various processes, namely Unified Process Metamodel, is presented. This universal solution serves the needs of modeling of modern decision-making systems used in a number of scientific areas. The metamodel uses solutions based on various knowledge engineering techniques. Four functional layers are selected under a new method, in each of them the specific types of components are defined. The particular attention to resource flow and control layer is devoted in the paper.","PeriodicalId":375852,"journal":{"name":"2018 5th International Conference on Systems and Informatics (ICSAI)","volume":"32 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"124456793","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 : 2018-11-01DOI: 10.1109/ICSAI.2018.8599479
Ruimei Zhang, S. Zhong, Deqiang Zeng
The exponential synchronization is studied in this paper for delayed memristive neural networks (MNNs). A new discontinuous adaptive control scheme is designed, which employs not only the proportional action but also the derivative action. Then, by adopting the adaptive control scheme and constructing an appropriate Lyapunov-Krasovskii functional (LKF), novel synchronization conditions are established. In the end, we use a numerical example to verify the effectiveness of the theory results.
{"title":"Adaptive Control for Exponential Synchronization of Delayed Memristive Neural Networks","authors":"Ruimei Zhang, S. Zhong, Deqiang Zeng","doi":"10.1109/ICSAI.2018.8599479","DOIUrl":"https://doi.org/10.1109/ICSAI.2018.8599479","url":null,"abstract":"The exponential synchronization is studied in this paper for delayed memristive neural networks (MNNs). A new discontinuous adaptive control scheme is designed, which employs not only the proportional action but also the derivative action. Then, by adopting the adaptive control scheme and constructing an appropriate Lyapunov-Krasovskii functional (LKF), novel synchronization conditions are established. In the end, we use a numerical example to verify the effectiveness of the theory results.","PeriodicalId":375852,"journal":{"name":"2018 5th International Conference on Systems and Informatics (ICSAI)","volume":"81 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"126319221","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 : 2018-11-01DOI: 10.1109/ICSAI.2018.8599496
Xinyu Lu, Lifang Wang, Zejun Jiang
HIV-1 protease cleavage site is critical for the design of HIV-1 protease inhibitors. Classification algorithms based on traditional machine learning are often used to deal with the prediction of HIV-1 protease cleavage sites. Unlike the classification algorithms of machine learning, the classification algorithms based on deep learning can extract the characteristics of the data well and get better performance. In this paper, HIV-1 protease cleavage site data is innovatively converted to One-hot data, and then two better classification models are proposed based on RNN and LSTM. At last, the experimental results are compared with the support vector machine algorithm and the random forest algorithm in traditional machine learning algorithm. The results show that the network structure based on deep learning designed in this paper can achieve higher accuracy than traditional algorithms after the HIV-1 protease cleavage site data is One-hot encoded, and the effects of RNN and LSTM are outstanding. Furthermore, the RNN-based classifier and LSTM-based classifier in this paper have much better Recall rate and F1-Measure than CNN and have high generalization ability.
{"title":"The Application of Deep Learning in the Prediction of HIV-1 Protease Cleavage Site","authors":"Xinyu Lu, Lifang Wang, Zejun Jiang","doi":"10.1109/ICSAI.2018.8599496","DOIUrl":"https://doi.org/10.1109/ICSAI.2018.8599496","url":null,"abstract":"HIV-1 protease cleavage site is critical for the design of HIV-1 protease inhibitors. Classification algorithms based on traditional machine learning are often used to deal with the prediction of HIV-1 protease cleavage sites. Unlike the classification algorithms of machine learning, the classification algorithms based on deep learning can extract the characteristics of the data well and get better performance. In this paper, HIV-1 protease cleavage site data is innovatively converted to One-hot data, and then two better classification models are proposed based on RNN and LSTM. At last, the experimental results are compared with the support vector machine algorithm and the random forest algorithm in traditional machine learning algorithm. The results show that the network structure based on deep learning designed in this paper can achieve higher accuracy than traditional algorithms after the HIV-1 protease cleavage site data is One-hot encoded, and the effects of RNN and LSTM are outstanding. Furthermore, the RNN-based classifier and LSTM-based classifier in this paper have much better Recall rate and F1-Measure than CNN and have high generalization ability.","PeriodicalId":375852,"journal":{"name":"2018 5th International Conference on Systems and Informatics (ICSAI)","volume":"55 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128137798","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}
With the development of information technology, it becomes more and more difficult to retrieve information from the internet for users. Question Answering (QA) is one of the methods to solve this problem. The users type natural language questions and get answers in QA systems. However, most QA systems only return a word or several words to the user, which is not friendly enough. The users are more willing to receive not only answers but also additional introductions or reasons. In this work, we propose a Nature Language Question Answering system which utilizes Seq2Seq model and Generative Adversarial Network (GAN) to generate answers with more information for users. To our best knowledge, this is the first work generating natural language answers in Question Answering domain. Our experiment results show NLQA can generate readable answers for users.
{"title":"Automatically Answering Questions With Nature Languages","authors":"Haitao Zheng, Jin-Yuan Chen, Zuo-You Fu, Zi-Han Xu, Cong-Zhi Zhao","doi":"10.1109/ICSAI.2018.8599337","DOIUrl":"https://doi.org/10.1109/ICSAI.2018.8599337","url":null,"abstract":"With the development of information technology, it becomes more and more difficult to retrieve information from the internet for users. Question Answering (QA) is one of the methods to solve this problem. The users type natural language questions and get answers in QA systems. However, most QA systems only return a word or several words to the user, which is not friendly enough. The users are more willing to receive not only answers but also additional introductions or reasons. In this work, we propose a Nature Language Question Answering system which utilizes Seq2Seq model and Generative Adversarial Network (GAN) to generate answers with more information for users. To our best knowledge, this is the first work generating natural language answers in Question Answering domain. Our experiment results show NLQA can generate readable answers for users.","PeriodicalId":375852,"journal":{"name":"2018 5th International Conference on Systems and Informatics (ICSAI)","volume":"108 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"121948140","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 : 2018-11-01DOI: 10.1109/ICSAI.2018.8599400
Zhihe Yang
In order to accurately predict the fracture fault of fan tooth belt, the NARIMA method is proposed in this paper. The method is based on ARIMA model, and effectively combines the run length stationary test method, differential stationary processing method, linear minimum variance prediction algorithm, etc.. The model is used to fit the time series of the fracture fault of fan tooth belt, and the model is used to predict the fracture fault of fan tooth belt. It is found that the NARIMA model can well fit the given time series, and the predicted values are in line with the actual situation and trend. The test results show the effectiveness of the proposed algorithm.
{"title":"A Prediction Algorithm For the Fan Tooth Belt Fracture Fault Based on Big Data","authors":"Zhihe Yang","doi":"10.1109/ICSAI.2018.8599400","DOIUrl":"https://doi.org/10.1109/ICSAI.2018.8599400","url":null,"abstract":"In order to accurately predict the fracture fault of fan tooth belt, the NARIMA method is proposed in this paper. The method is based on ARIMA model, and effectively combines the run length stationary test method, differential stationary processing method, linear minimum variance prediction algorithm, etc.. The model is used to fit the time series of the fracture fault of fan tooth belt, and the model is used to predict the fracture fault of fan tooth belt. It is found that the NARIMA model can well fit the given time series, and the predicted values are in line with the actual situation and trend. The test results show the effectiveness of the proposed algorithm.","PeriodicalId":375852,"journal":{"name":"2018 5th International Conference on Systems and Informatics (ICSAI)","volume":"120 2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132351771","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 : 2018-11-01DOI: 10.1109/ICSAI.2018.8599317
Dongliang Guo
A design of low-density parity-check (LDPC) code via the masking technology and progressive optimization is proposed. The code generated by this method has lower computational complexity, and their parity-check matrices can effectively refrain from the girth −4 phenomenon. The proposed method has the superiorities such as better girth-length characteristic and more flexible trim in the length and rate of the code. Experiment results indicate that the error-correction performance of the new code should be better than or as good as the capability of the code which is constructed without been masking. Futhermore, under the condition of short code length, the performance is better than that of LDPC codes via the randomized construction methods.
{"title":"LDPC Code Design via Masking Technology and Progressive Optimization","authors":"Dongliang Guo","doi":"10.1109/ICSAI.2018.8599317","DOIUrl":"https://doi.org/10.1109/ICSAI.2018.8599317","url":null,"abstract":"A design of low-density parity-check (LDPC) code via the masking technology and progressive optimization is proposed. The code generated by this method has lower computational complexity, and their parity-check matrices can effectively refrain from the girth −4 phenomenon. The proposed method has the superiorities such as better girth-length characteristic and more flexible trim in the length and rate of the code. Experiment results indicate that the error-correction performance of the new code should be better than or as good as the capability of the code which is constructed without been masking. Futhermore, under the condition of short code length, the performance is better than that of LDPC codes via the randomized construction methods.","PeriodicalId":375852,"journal":{"name":"2018 5th International Conference on Systems and Informatics (ICSAI)","volume":"20 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"129977824","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 : 2018-11-01DOI: 10.1109/ICSAI.2018.8599372
Zheng Hu, Yongping Li, Zhiyong Yang
Rectified linear unit (ReLU) is a widely used activation function in artificial neural networks, it is considered to be an efficient active function benefit from its simplicity and nonlinearity. However, ReLU’s derivative for negative inputs is zero, which can make some ReLUs inactive for essentially all inputs during the training. There are several ReLU variations for solving this problem. Comparing with ReLU, they are slightly different in form, and bring other drawbacks like more expensive in computation. In this study, pseudo derivatives were tried replacing original derivative of ReLU while ReLU itself was unchanged. The pseudo derivative was designed to alleviate the zero derivative problem and be consistent with original derivative in general. Experiments showed using pseudo derivative ReLU (PD-ReLU) could obviously improve AlexNet (a typical convolutional neural network model) in CIFAR-10 and CIFAR-100 tests. Furthermore, some empirical criteria for designing such pseudo derivatives were proposed.
{"title":"Improving Convolutional Neural Network Using Pseudo Derivative ReLU","authors":"Zheng Hu, Yongping Li, Zhiyong Yang","doi":"10.1109/ICSAI.2018.8599372","DOIUrl":"https://doi.org/10.1109/ICSAI.2018.8599372","url":null,"abstract":"Rectified linear unit (ReLU) is a widely used activation function in artificial neural networks, it is considered to be an efficient active function benefit from its simplicity and nonlinearity. However, ReLU’s derivative for negative inputs is zero, which can make some ReLUs inactive for essentially all inputs during the training. There are several ReLU variations for solving this problem. Comparing with ReLU, they are slightly different in form, and bring other drawbacks like more expensive in computation. In this study, pseudo derivatives were tried replacing original derivative of ReLU while ReLU itself was unchanged. The pseudo derivative was designed to alleviate the zero derivative problem and be consistent with original derivative in general. Experiments showed using pseudo derivative ReLU (PD-ReLU) could obviously improve AlexNet (a typical convolutional neural network model) in CIFAR-10 and CIFAR-100 tests. Furthermore, some empirical criteria for designing such pseudo derivatives were proposed.","PeriodicalId":375852,"journal":{"name":"2018 5th International Conference on Systems and Informatics (ICSAI)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-11-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"130104893","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：