An ensemble kernel classifier with immune clonal selection algorithm for automatic discriminant of primary open-angle glaucoma

The least-cost multicast routing with delay constrained is an NP-Complete problems, to deal with which, a Multi-Agent Immune Clonal Selection Algorithm based multicast routing (MAICSA) is proposed in this paper. MAICSA combines the characteristic of Multi-Agent with the search strategy of Immune Clonal Selection Algorithm. To compare with the conventional Genetic Algorithm (GA), MAICSA overcomes the most serious drawbacks, such as slow convergence rate and “prematurity”. The experimental results show that MAICSA has faster astringency and higher precision than traditional GA, MAGA (Multi-Agent multicast routing based on Genetic Algorithm) and MAIA (Multi-Agent multicast routing based on Immune Algorithm).

In this paper, the mixed-integer nonlinear programming model is established for hybrid flow-shop scheduling problem (HFSP) with the minimum of makespan as the objective function. In order to reduce the computational complexity, immune clonal selection algorithm (ICSA) is applied to HFSP. The definitions of antibody affinity, comparability and density are given in detail. To improve the ability of global optimization for ICSA, mutliclone operator (mutation, crossover and selection) and grouping strategy are employed. The simulation results indicate that ICSA can obtain preferable effect for the solution to HFSP.

One of the main implementation disadvantages of a multicarrier communication system is the possibly high peak to average power ratio of the transmitted signals which cause the requirement of highly cost linear amplifiers with large dynamic range. One proposed solution is given by Haiming Wang [1] which is based on the algorithm of Hopfield neural network (HNN). Also, in our previous work [2],we demonstrated the solution based on the immune clonal selection algorithm (ICSA) which has a better performance than [1]. However, a important disadvantage of the ICSA is the need of high number of iteration. In this paper, we will show a hybrid solution which adopts both the concept of ICSA and HNN. According to the simulation results, this solution maintained the good performance of PAPR reduction, meanwhile, the number of iteration is significantly reduced.

The Sacramento model is widely utilized in hydrological forecast, of which the accuracy and performance are primarily determined by the model parameters, indicating the key role of parameter estimation. This paper presents a multi-step parameter estimation method, which divides the parameter estimation of Sacramento model into three steps and realizes optimization step by step. We firstly use the immune clonal selection algorithm (ICSA) to solve the non-liner objective function of parameter estimation, and compare the parameter calibration result of ideal artificial data with Shuffled Complex Evolution (SCE-UA), Parallel Genetic Algorithm (PGA), and Serial Master-slaver Swarms Shuffling Evolution Algorithm Based on Particle Swarms Optimization (SMSE-PSO). The comparison result shows that ICSA has the best convergence, efficiency and precision. Then we apply ICSA to the parameter estimation of single-step and multi-step Sacramento model and simulate 32 floods based on application examples of Dongyang and Tantou river basins for validation. It is clearly shown that the results of multi-step method based on ICSA show higher accuracy and 100% qualified rate, indicating its higher precision and reliability, which has great potential to improve Sacramento model and hydrological forecast.

PurposeThe extensive applications of the industrial robots have made the optimization of assembly lines more complicated. The purpose of this paper is to develop a balancing method of both workstation time and station area to improve the efficiency and productivity of the robotic assembly lines. A tradeoff was made between two conflicting objective functions, minimizing the number of workstations and minimizing the area of each workstation.Design/methodology/approachThis research proposes an optimal method for balancing robotic assembly lines with space consideration and reducing robot changeover and area for tools and fixtures to further minimize assembly line area and cycle time. Due to the NP-hard nature of the considered problem, an improved multi-objective immune clonal selection algorithm is proposed to solve this constrained multi-objective optimization problem, and a special coding scheme is designed for the problem. To enhance the performance of the algorithm, several strategies including elite strategy and global search are introduced.FindingsA set of instances of different problem scales are optimized and the results are compared with two other high-performing multi-objective algorithms to evaluate the efficiency and superiority of the proposed algorithm. It is found that the proposed method can efficiently solve the real-world size case of time and space robotic assembly line balancing problems.Originality/valueFor the first time in the robotic assembly line balancing problems, an assignment-based tool area and a sequence-based changeover time are took into consideration. Furthermore, a mathematical model with bi-objective functions of minimizing the number of workstations and area of each station was developed. To solve the proposed problem, an improved multi-objective immune clonal selection algorithm was proposed and a special coding scheme is designed.

Water is an important natural resource in our daily life. Water resources play a major role in sustainable development. Surveying and continuous monitoring of water bodies in a vast area is a tedious and timeconsuming task by conventional surveying techniques. Keeping track of changes in the areas of waterbodies due to urbanization and industrialization is very important. The present study aims at determining the spatiotemporal changes in the water bodies from 2008 to 2018 in the drought-hit area of Mahbubnagar, Telangana using satellite-derived spectral index namely,Modified Normalized Difference Water Index (MNDWI) and Support Vector Machine (SVM) technique using Resourcesat-1 Data. Extraction of water bodies using MNDWI and SVM can be done very easily with very little effort as compared to conventional surveying techniques. Using the MNDWI technique,the total area of waterbodies is 322.64 square kilometers in the year 2008 and 362.61 square kilometers in the year 2018. Using the SVM technique, the area of waterbodies is 368.17 square kilometers in the year 2008 and 395.72 in the year 2018. The kappa index values of 2018 by MNDWI technique and SVM technique are 0.86 and 0.91 respectively. The kappa index values of 2008 by MNDWI technique and SVM technique are 0.85 and 0.89 respectively. It is observed that the SVM technique is more accurate than the MNDWI technique. The area of water bodies has increased during the study period mainly due to increased awareness among the public on the importance of water and also initiatives taken by the Government of Telangana, like Mission Kakatiya for the revival of water bodies in the study area.

Both empty and heavy car flows should be taken into consideration in the study of the Optimization Model of Car Flow Organization, in the Loading End of Heavy Haul Railway. Firstly, through the analysis of empty and heavy car flow organization in the semi-closed heavy haul railway loading end, comprehensive optimization model based on ability of station and railway can be established in order to obtain car flow organization scheme with minimum residence time of both empty and heavy cars in the loading end. Secondly, immune clone algorithm is applied to solve the model in which the objective function is regarded as antibody and the constrained condition is considered as antigen to calculate the affinity, further to conduct the mutation, based on information entropy. Mutation is conducted according to the affinity. In the solving process, reproduction can be realized with the use of immune clone selection algorithm. Besides, premature convergence can be prevented by using the antibody concentration which can control the population size. Finally, simulation results show that the average search time of the proposed algorithm is reduced by 36 and 51% respectively compared with the GA and PSO.

Cancer is the most common disease to threaten human health. The ability to screen individuals with malignant tumours with only a blood sample would be greatly advantageous to early diagnosis and intervention. This study explores the possibility of discriminating between cancer patients and normal subjects with serum surface-enhanced Raman spectroscopy (SERS) and a support vector machine (SVM) through a peripheral blood sample. A total of 130 blood samples were obtained from patients with liver cancer, colonic cancer, esophageal cancer, nasopharyngeal cancer, gastric cancer, as well as 113 blood samples from normal volunteers. Several diagnostic models were built with the serum SERS spectra using SVM and principal component analysis (PCA) techniques. The results show that a diagnostic accuracy of 85.5% is acquired with a PCA algorithm, while a diagnostic accuracy of 95.8% is obtained using radial basis function (RBF), PCA–SVM methods. The results prove that a RBF kernel PCA–SVM technique is superior to PCA and conventional SVM (C-SVM) algorithms in classification serum SERS spectra. The study demonstrates that serum SERS, in combination with SVM techniques, has great potential for screening cancerous patients with any solid malignant tumour through a peripheral blood sample.

Aimed at the nonlinear properties of analog circuits features , kernel Principal Component Analysis(KPCA) combined with Support Vector Machine(SVM) based diagnostic method for faults of analog circuit was proposed in this paper. Here, uses KPCA to reduce the irrelevant features. The KPCA method is used to extract the initial features. The SVM is then applied to the circuit after features extraction. It can be used as a new idea for Fault Diagnosis of Analog Circuits. To better verify the superiority of the proposed method, KPCA+SVM classification results were compared with the results of PCA+SVM. It is concluded that KPCA+SVM classifier achieved a better performance than PCA+SVM in the terms of the accuracy.

In this paper, a novel HVAC fan machinery fault recognition method combining kernel principal component analysis (KPCA) and support vector machine (SVM) is proposed. KPCA is an improved PCA, which possesses the property of extracting optimal features by adopting a nonlinear kernel function method. Support vector machine (SVM) is a novel approach based on statistical learning theory, which has emerged for feature identification and classification. An integrated method is applied for HVAC fan machinery status monitoring and fault diagnosis, which combines KPCA for fault feature extraction and multiple SVMs (MSVMs) for identification of different fault sources. The experimental results show that KPCA based on LS-SVM has a higher correct recognition rate, and a faster computational speed.

Automated classification of magnetic resonance brain images (MRIs) is a hot topic in the field of medical and biomedical imaging. Various methods have been suggested recently to improve this technology. In this paper, to reduce the complexity involved in the medical images and to ameliorate the classification of MRIs, a novel 3D magnetic resonance (MR) brain image classifier using kernel principal component analysis (KPCA) and support vector machines (SVMs) is proposed. Experiments are carried out using A deep multiple kernel SVM (DMK-SVM) and a regular SVM. An algorithm entitled SVM–KPCA is put forward. Its main task is to classify a brain MRI as a normal brain image or as a pathological brain image. This algorithm, firstly, adopts the discrete wavelet transform technique to extract features from images. Secondly, KPCA is applied to decrease the dimensionality of features. SVM is then applied to the reduced data. A K-fold cross-validation strategy is used to avoid overfitting and to ameliorate the generalization of the SVM–KPCA algorithm. Three databases are used to validate the suggested SVM–KPCA method. Three conclusions are obtained from this work. First, KPCA is highly efficient in increasing the classifier’s performance compared with similar algorithms working on the proposed database. Second, the SVM–KPCA algorithm performs well in differentiating between two classes of medical images. Third, the approach is robust and might be utilized for other MRIs. This proposes a significant role for computer aided diagnosis analysis systems used for clinical practice.

This study investigates stock market portfolio construction that is an interesting and important research in the areas of investment and applications, as it allows more profits and returns at lower risk rate with effective portfolios. To construct accurate prediction and thus good portfolio construction, various methods have been tried, among which the machine learning methods have drawn attention and been developed. This paper introduces a machine learning method of Support Vector Machine (SVM) to construct a stocks portfolio by selecting well-performing firms in Casablanca Stock Exchange (Casablanca S.E). This model can perform a nonlinear classification. However, the accuracy of SVM classification is particularly sensitive to the quality of data. To insure a better performance of the model, we bring the Kernel Principal Component Analysis (KPCA) into SVM to extract the low-dimensional and efficient feature information. As empirical results show, based on SVM, within KPCA, the stock selection model outperforms the MASI index.

Machine learning and pattern recognition recently become a hot topic in computing world. This is due to the fast-growing of resources as well as techniques that make it easier to solve machine learning and pattern recognition problems. Problems that require machine learning solutions may be very simple for humans but actually can be very complex for machines to solve them. Face recognition is amongst those problems. Almost all human can easily recognize others without require specific knowledge to do it, different from machines which require its. This paper discussed face recognition task using machine learning strategies which involved Kernel Principal Component Analysis (KPCA) and Support Vector Machine (SVM) to identify person. KPCA extracted features from 2D image input and produced the important features of an image input. The extracted face features are recognized by SVM by classifying human face according to their stored identity in a database. SVM, which was basically a binary classifier worked by using one-against-one strategy to compare the face feature vector of a single test image to the stored face image in a face image database. Experiment results on grayscale images with size 92x112 pixels gave 96.25% of accuracy rate. Hence, KPCA and SVM for face recognition is a robust machine learning method.

Aimed at the non-stationary characteristics of rolling bearing vibration signal,a fault diagnosis method was proposed based on local-wave method and KPCA(kernel principal component analysis)-LSSVM(least squares support vector machine).Firstly,local wave decomposition was used to decompose rolling bearing vibration signal into several intrinsic mode function(IMF),whose feature energy,singular values and AR model parameters were computed as initial feature vectors.Secondly,ini tial feature vectors were mapped into a higher-dimensional space with KPCA,and the kemel principal components were extracted as new feature vectors,which used as the input of LSSVM for fault classification.The experimental results show the KPCA-LSSVM method improves LSSVM's classification performance by KPCA obtaining additional discriminative information,and has better generalization than direct LSSVM method,and can identify rolling bearing fault patterns more accurately.

The problem with development of the support vector machine (SVM) classifiers using modified particle swarm optimization (PSO) algorithm and their ensembles has been considered. Solving this problem would allow fulfilling the high-precision data classification, especially Big Data classification, with the acceptable time expenditures. The modified PSO algorithm conducts a simultaneous search of the type of kernel functions, the parameters of the kernel function and the value of the regularization parameter for the SVM classifier. The idea of particles' «regeneration» served as the basis for the modified PSO algorithm. In the implementation of this algorithm, some particles change the type of their kernel function to the one which corresponds to the particle with the best value of the classification accuracy. The offered PSO algorithm allows reducing the time expenditures for the developed SVM classifiers, which is very important for Big Data classification problem. In most cases such SVM classifier provides the high quality of data classification. In exceptional cases the SVM ensembles based on the decorrelation maximization algorithm for the different strategies of the decision-making on the data classification and the majority vote rule can be used. Also, the two-level SVM classifier has been offered. This classifier works as the group of the SVM classifiers at the first level and as the SVM classifier on the base of the modified PSO algorithm at the second level. The results of experimental studies confirm the efficiency of the offered approaches for Big Data classification.