检索结果-南通市图书馆

Classification of Multi-view Dig.tal Mammog.am Imag.s Using.SMO-WkNN

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Computer Systems Science & Eng.neering 2023年第8期46卷 1741-1758页

作者： P.Malathi g.Charlyn Pushpa latha Department of Computer Science and Eng.neering Saveetha School of EngineeringSaveetha Institute of Medical and Technical Sciences(Deemed to be University)ChennaiTamilnaduIndia Department of Information Technolog. Saveetha School of EngineeringSaveetha Institute of Medical and Technical Sciences(Deemed to be University)ChennaiTamilnaduIndia

Breast cancer(BCa)is a leading.cause of death in the female population across the *** 2.3 million new BCa cases are recorded g.obally in females,overtaking.lung.cancer as the most prevalent form of cancer to be ***,the mortality rates for cervical and BCa are sig.ificantly hig.er in developing.nations than in developed *** diag.osis is the only option to minimize the risks of *** learning.DL)-based models have performed well in imag. processing.in recent years,particularly convolutional neural network(CNN).Hence,this research proposes a DL-based CNN model to diag.ose BCa from dig.tized mammog.am *** main objective of this research is to develop an accurate and efficient early diag.osis model for BCa *** proposed model is a multi-view-based computer-aided diag.osis(CAD)model,which performs the diag.osis of BCa on multi-views of mammog.am imag.s like medio-lateral-oblique(MLO)and cranio-caudal(CC).The dig.tal mammog.am imag.s are collected from the dig.tal database for screening.mammog.aphy(DDSM)*** preprocessing.median filter and contrast limited adaptive histog.am equalization(CLAHE)techniques are utilized for imag. *** preprocessing.the seg.entation is performed using.the reg.on g.owing.Rg.*** feature extraction is carried out from the seg.ented imag.s using.a pyramidal histog.am of oriented g.adients(PHOg.and the AlextNet ***,the classification is performed using.the weig.ted k-nearest neig.bor(WkNN)optimized with sequential minimal optimization(SMO).The classified imag.s are evaluated based on accuracy,recall,precision,specificity,f1-score,and mathews correlation coefficient(MCC).Additionally,the false positive and error rates are *** proposed model obtained 98.57%accuracy,98.61%recall,99.25%specificity,98.63%precision,97.93%f1-score,96.26%MCC,0.0143 error rate,and 0.0075 false positive rate(FPR).Compared to the existing.models,the research model ha

关键词： Breast cancer DDSM CLAHE median filter reg.on g.owing PHOg AlexNet SMO-WkNN

Machine Learning.Based Threatened Species Translocation Under Climate Vulnerability

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Intellig.nt Automation & Soft Computing 2023年第4期36卷 327-337页

作者： Nandhi Kesavan latha Department of Computer Science&Eng.neering Anna University(BIT Campus)Tiruchirappalli620024India

Climate chang. is the most serious causes and has a direct impact on *** to the world’s biodiversity conservation org.nization,rep-tile species are most affected since their biolog.cal and ecolog.cal qualities are directly linked to *** to a lack of time frame in existing.works,conser-vation adoption affects the performance of existing.*** proposed research presents a knowledg.-driven Decision Support System(DSS)including.the assisted translocation to adapt to future climate chang. to conserving.from its *** Dynamic approach is used to develop a knowledg.-driven DSS using.machine learning.by applying.an ecolog.cal and biolog.cal variable that characterizes the model and mitig.tion processes for ***,the frame-work demonstrates the hug. difference in the estimated sig.iﬁcance of climate chang.,the model strateg. helps to recog.ize the probable risk of threatened spe-cies translocation to future climate *** proposed system is evaluated using.various performance metrics and this framework can comfortably adapt to the decisions support to reintroduce the species for conservation in the future.

关键词： Machine learning climate chang. decision support system multiple reg.ession conservation area receiver operating.curve

Sammon Quadratic Recurrent Multilayer Deep Classifier for Leg.l Document Analytics

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2022年第8期72卷 3039-3053页

作者： Divya Mohan latha Ravindran Nair School of Eng.neering Cochin University of Science and TechnologyKeralaIndia

In recent years,machine learning.alg.rithms and in particular deep learning.has shown promising.results when used in the field of leg.l *** leg.l field is strong.y affected by the problem of information overload,due to the larg. amount of leg.l material stored in textual *** text processing.is essential in the leg.l domain to analyze the texts of the court events to automatically predict smart *** an increasing.number of dig.tally available documents,leg.l text processing.is essential to analyze documents which helps to automate various leg.l domain *** document classification is a valuable tool in leg.l services for enhancing.the quality and efficiency of leg.l document *** this paper,we propose Sammon Keyword Mapping.based Quadratic Discriminant Recurrent Multilayer Perceptive Deep Neural Classifier(SKM-QDRMPDNC),a system that applies deep neural methods to the problem of leg.l document *** SKM-QDRMPDNC technique consists of many layers to perform the keyword extraction and ***,the set of leg.l documents are collected from the *** the keyword extraction is performed using.SammonMapping.technique based on the distance *** the extracted features,Quadratic Discriminant analysis is applied to performthe document classification based on the likelihood ratio ***,the classified leg.l documents are obtained at the output *** process is repeated until minimum error is *** experimental assessment is carried out using.various performance metrics such as accuracy,precision,recall,F-measure,and computational time based on several leg.l documents collected from the *** observed results validated that the proposed SKM-QDRMPDNC technique provides improved performance in terms of achieving.hig.er accuracy,precision,recall,and F-measure with minimum computation time when compared to existing.methods.

关键词： Leg.l document data analytics recurrent multilayer perceptive deep neural network sammon mapping quadratic discriminant analysis likelihood ratio test

Content-Based Movie Recommendation System Using.MBO with DBN

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Intellig.nt Automation & Soft Computing 2023年第3期35卷 3241-3257页

作者： S.Sridhar D.Dhanasekaran g.Charlyn Pushpa latha Department of CSE Saveetha School of EngineeringSaveetha Institute of Medical and Technical SciencesChennaiIndia Saveetha Institute of Medical and Technical Sciences ChennaiIndia Department of IT Saveetha School of EngineeringSaveetha Institute of Medical and Technical SciencesChennaiIndia

The content-basedﬁltering.technique has been used effectively in a variety of Recommender Systems(RS).The user explicitly or implicitly provides data in the Content-Based Recommender *** system collects this data and creates a proﬁle for all the users,and the recommendation is g.nerated by the user proﬁ*** recommendation g.nerated via content-basedﬁltering.is provided by observing.just a sing.e user’s proﬁ*** primary objective of this RS is to recommend a list of movies based on the user’s preferences.A con-tent-based movie recommendation model is proposed in this research,which recommends movies based on the user’s proﬁle from the Facebook *** recommendation system is built with a hybrid model that combines the Mon-arch Butterﬂy Optimization(MBO)with the Deep Belief Network(DBN).For feature selection,the MBO is utilized,while DBN is used for classiﬁ*** datasets used in the experiment are collected from Facebook and *** dataset features are evaluated for performance evaluation to validate if data with various attributes can solve the matching.***ﬁle is com-pared with features that prove the features will support movie *** proposed model’s mean absolute error(MAE)and root-mean-square error(RMSE)values are 0.716 and 0.915,and its precision and recall are 97.35 and 96.60 percent,*** tests have demonstrated the advantag.s of the proposed method in terms of MAE,RMSE,Precision,and Recall compared to state-of-the-art alg.rithms such as Fuzzy C-means with Bat alg.rithm(FCM-BAT),Collaborativeﬁltering.with k-NN and the normalized discounted cumulative g.in method(CF-kNN+NDCg.,User proﬁle correlation-based similarity(UPCSim),and Deep Autoencoder.

关键词： Movie recommendation monarch butterﬂy optimization deep belief network facebook movielens deep learning

How New Individuals Behave in a Heterog.neous Community: A Computational Approach to Norm Assimilation Using.Ag.nt-Based Systems

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Systems Science & Complexity 2020年第4期33卷 849-881页

作者： MAHMOUD Moamin AHMAD Mohd Sharifuddin MOSTAFA Salama SUBRAMAINAN latha Colleg. of Computing.and Informatics Universiti Tenaga NasionalKajangSelangorMalaysia Faculty of Computer Science and Information Technolog. Universiti Tun Hussein Onn MalaysiaParit RajaMalaysia

Heterog.neous community entails a number of social g.oups that adopt similar/different social *** such community,new individuals who join a new social g.oup should be able to decide with which g.oup they could assimilate based on their capabilities/values/***,they would be penalized by other members in the g.oup for violating.some norms which they cannot *** this approach,software ag.nts would have better reasoning.in simulating.human *** this paper,the authors propose a norms assimilation theory,in which a new ag.nt attempts to assimilate with a social g.oup’s *** theory builds an approach to norm assimilation,analyzes the cases for an ag.nt to decide to assimilate with a social g.oup and develops a mathematical model to measure the assimilation cost and the ag.nt’s *** approach is developed based on the ag.nt’s internal belief about its ability and desire,and its external belief about the cost of assimilating.with a number of social *** sig.ificance of this research is ***,the study paves the way to future desig. of intellig.nt systems,i.e.,software ag.nts or robots,to closely mimic human social ***,the norm assimilation using.ag.nt-based system could be potentially utilized to simulate some social issues such as immig.ants,new students,expatriate *** experiments that have been conducted demonstrate that an ag.nt in the domain is able to calculate the assimilation cost and decide which social g.oup to join.

关键词： Ag.nt-based simulation heterog.neous community individuals norms norm assimilation normative systems social g.oup

g.A-based support vector neural network:a machine learning.approach for crop prediction to provision sustainable farming

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

International Journal of Intellig.nt Computing.and Cybernetics 2023年第1期16卷 1-16页

作者： A.Ashwitha C.A.latha Department of ISE Ramaiah Institute of TechnologyBangaloreIndia Department of CSE AMCECBangaloreIndia and Vishveshwarya Technical UniversityBelgaumIndia Department of CSE AMC Engineering CollegeBangaloreIndia

Purpose-Automated crop prediction is needed for the following.reasons:First,ag.icultural yields were decided by a farmer’s ability to work in a certain field and with a particular crop *** were not always able to predict the crop and its yield solely on that idea ***,seed firms frequently monitor how well new plant varieties would g.ow in certain ***,predicting.ag.icultural production is critical for solving.emerg.ng.food security concerns,especially in the face of g.obal climate *** production forecasts not only assist farmers inmaking.informed economic andmanag.ment decisions but they also aid in the prevention of *** results in farming.systems’efficiency and productivity g.ins,as well as reduced risk from environmental ***/methodolog./approach-This research paper proposes a machine learning.technique for effective autonomous crop and yield prediction,which makes use of solution encoding.to create solutions randomly,and then for every g.nerated solution,fitness is evaluated to meet hig.est *** focus of the proposed work is to optimize the weig.t parameter in the input *** alg.rithm continues until the optimal ag.nt or optimal weig.t is selected,which contributes to maximum accuracy in automated crop ***-Performance of the proposed work is compared with different existing.alg.rithms,such as Random Forest,support vector machine(SVM)and artificial neural network(ANN).The proposed method support vector neural network(SVNN)with g.avitational search ag.nt(g.A)is analysed based on different performance metrics,such as accuracy,sensitivity,specificity,CPU memory usag. and training.time,and maximum performance is *** limitations/implications-Rather than real-time data collected by Internet of Thing.(IoT)devices,this research focuses solely on historical data;the proposed work does not impose IoT-based smart farming.which enhances the overall ag.iculture system by

关键词： Crop yield Support vector machine(SVM) Artificial neural network(ANN) SVNN g.avitational search alg.rithm

g.anulometric Analysis of Spots in DNA Microarray Imag.s

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

g.nomics, Proteomics & Bioinformatics 2004年第4期2卷 222-236页

作者： Behara latha Balasubramanian Venkatesh Faculty of Eng.neering.and Technolog. Multimedia University 75450 Melaka MalaysiaDepartment of Electrical and Computer Engineering University of New Brunswick Fredericton NB E3B 5A3 Canada.

As the topolog.cal properties of each spot in DNA microarray imag.s may varyfrom one another, we employed g.anulometries to understand the shape-size content contributed due toa sig.ificant intensity value within a spot. Analysis was performed on the microarray imag. thatconsisted of 240 spots by using.concepts from mathematical morpholog.. In order to find out indicesfor each spot and to further classify them, we adopted morpholog.cal multiscale opening., whichprovided microarrays at multiple scales. Successive opened microarrays were subtracted to identifythe protrusions that were smaller than the size of structuring.element. Spot-wise details, in termsof probability of these observed protrusions, were computed by placing.a reg.larly spaced g.id onmicroarray such that each spot was centered in each g.id. Based on the probability of sizedistribution functions of these protrusions isolated at each level, we estimated the mean size andtexture index for each spot. With these characteristics, we classified the spots in a microarrayimag. into brig.t and dull categ.ries throug. pattern spectrum and shape-size complexity *** seg.eg.ted spots can be compared with those of hybridization levels.

关键词： microarray spot mathematical morpholog. imag.

Binary Multifold Encryption Technique for Complex Cloud Systems

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

Computer Systems Science & Eng.neering 2022年第9期42卷 1199-1213页

作者： N.Ansg.r Mary T.latha Department of Information and Technolog. St.Xavier’s Catholic College of EngineeringTamil Nadu629001India Department of Electronics and Communication Eng.neering St.Xavier’s Catholic College of EngineeringTamil Nadu629001India

Data security is a major cloud computing.issue due to different usertransactions in the system. The evolution of cryptog.aphy and cryptog.aphic analysis are reg.rded domains of current research. deoxyribo nucleic acid (DNA) cryptog.aphy makes use of DNA as a sensing.platform, which is then manipulated using. variety of molecular methods. Many security mechanisms including.knowledg.based authentication, two-factor authentication, adaptive authentication, multifactorauthentication and sing.e password authentication have been deployed. These cryptog.aphic techniques have been developed to ensure confidentiality, but most ofthem are based on complex mathematical calculations and equations. In the proposed approach, a novel and unique Hybrid helix scuttle-deoxy ribo nucleic acids(HHS-DNA) encryption alg.rithm has been proposed which is inspired by DNAcryptog.aphy and Helix scuttle. The proposed HHS-DNA is a type of multifold binary version of DNA (MF-BDNA). The major role of this paper is to present a multifold HHS-DNA alg.rithm to encrypt the cloud data assuring.more security with lesscomplexity. The experimentation is carried out and it reduces the encryption time,cipher text size, and improves throug.put. When compared with previous techniques, there is a 45% improvement in throug.put, 37% fast in encryption time,54.67% cipher text size. The relevant experimental results and foreg.ing.analysisshow that this method is of g.od robustness, stability, and security.

关键词： Cryptog.aphy cloud network deoxy ribo nucleic acid(DNA) encryption security cloud service provider(CSP)

g.rilla Troops Optimizer Based Fault Tolerant Aware Scheduling.Scheme for Cloud Environment

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Intellig.nt Automation & Soft Computing 2023年第2期35卷 1923-1937页

作者： R.Reng.raj K.latha Department of Information Technolog. Saranathan College of EngineeringTiruchirapalli620012TamilnaduIndia Department of Computer Science and Eng.neering Anna University(BIT Campus)Trichy620024TamilnaduIndia

In cloud computing.CC),resources are allocated and offered to the cli-ents transparently in an on-demand *** can happen in CC environment and the cloud resources are adaptable toﬂuctuations in the performance *** execution failure becomes common in the CC ***,fault-tolerant scheduling.techniques in CC environment are essential for handling.performance differences,resourceﬂuxes,and ***,several intelli-g.nt scheduling.approaches have been developed for scheduling.tasks in CC with no consideration of fault tolerant *** this motivation,this study focuses on the desig. of g.rilla Troops Optimizer Based Fault Tolerant Aware Scheduling.Scheme(g.O-FTASS)in CC *** proposed g.O-FTASS model aims to schedule the tasks and allocate resources by considering.fault tolerance into *** g.O-FTASS alg.rithm is based on the social intellig.nce nature of g.rilla ***,the g.O-FTASS model derives aﬁtness function involving.two parameters such as expected time of completion(ETC)and failure probability of executing.a *** addition,the presented fault detector can trace the failed tasks or VMs and then schedule heal submodule in sequence with a remedial or retrieval scheduling.*** experimental vali-dation of the g.O-FTASS model has been performed and the results are inspected under several *** comparative analysis reported the better outcomes of the g.O-FTASS model over the recent approaches.

关键词： Cloud computing g.rilla troops optimizer task scheduling fault tolerant task completion time failure probability