Faculty Publications
Permanent URI for this communityhttps://idr.nitk.ac.in/handle/123456789/18736
Publications by NITK Faculty
Browse
4 results
Search Results
Item Genetic algorithm based support vector machine regression in predicting wave transmission of horizontally interlaced multi-layer moored floating pipe breakwater(Elsevier Ltd, 2012) Patil, S.G.; Mandal, S.; Hegde, A.V.Planning and design of coastal protection works like floating pipe breakwater require information about the performance characteristics of the structure in reducing the wave energy. Several researchers have carried out analytical and numerical studies on floating breakwaters in the past but failed to give a simple mathematical model to predict the wave transmission through floating breakwaters by considering all the boundary conditions. Computational intelligence techniques, such as, Artificial Neural Networks (ANN), fuzzy logic, genetic programming and Support Vector Machine (SVM) are successfully used to solve complex problems. In the present paper, a hybrid Genetic Algorithm Tuned Support Vector Machine Regression (GA-SVMR) model is developed to predict wave transmission of horizontally interlaced multilayer moored floating pipe breakwater (HIMMFPB). Furthermore, optimal SVM and kernel parameters of GA-SVMR models are determined by genetic algorithm. The GA-SVMR model is trained on the data set obtained from experimental wave transmission of HIMMFPB using regular wave flume at Marine Structure Laboratory, National Institute of Technology, Karnataka, Surathkal, Mangalore, India. The results are compared with ANN and Adaptive Neuro-Fuzzy Inference System (ANFIS) models in terms of correlation coefficient, root mean square error and scatter index. Performance of GA-SVMR is found to be reliably superior. b-spline kernel function performs better than other kernel functions for the given set of data. © 2011 Elsevier Ltd. All rights reserved.Item Particle Swarm Optimization based support vector machine for damage level prediction of non-reshaped berm breakwater(Elsevier Ltd, 2015) Narayana, N.; Mandal, S.; Rao, S.; Patil, S.G.The damage analysis of coastal structure is very much essential for better and safe design of the structure. In the past, several researchers have carried out physical model studies on non-reshaped berm breakwaters, but failed to give a simple mathematical model to predict damage level for non-reshaped berm breakwaters by considering all the boundary conditions. This is due to the complexity and non-linearity associated with design parameters and damage level determination of non-reshaped berm breakwater. Soft computing tools like Artificial Neural Network, Fuzzy Logic, Support Vector Machine (SVM), etc, are successfully used to solve complex problems. In the present study, SVM and hybrid of Particle Swarm Optimization (PSO) with SVM (PSO-SVM) are developed to predict damage level of non-reshaped berm breakwaters. Optimal kernel parameters of PSO-SVM are determined by PSO algorithm. Both the models are trained on the data set obtained from experiments carried out in Marine Structures Laboratory, Department of Applied Mechanics and Hydraulics, National Institute of Technology Karnataka, Surathkal, India. Results of both models are compared in terms of statistical measures, such as correlation coefficient, root mean square error and scatter index. The PSO-SVM model with polynomial kernel function outperformed other SVM models. © 2014 Elsevier B.V.Item Application of an evolutionary technique (PSO–SVM) and ANFIS in clear-water scour depth prediction around bridge piers(Springer London, 2019) Marulasiddappa, B.M.; Rao, M.; Mandal, S.The mechanism of the local scour around bridge pier is so complicated that it is hard to predict the scour accurately using a traditional method frequently by considering all the governing variables and boundary conditions. The present study aims to investigate the application of different hybrid soft computing algorithms, such as particle swarm optimization (PSO)-tuned support vector machine (SVM) and a hybrid artificial neural network-based fuzzy inference system to predict the scour depth around different shapes of the pier using experimental data. The important independent input parameters used in developing the soft computing models are sediment particle size, a velocity of the flow and the time taken in the prediction of the scour depth around the bridge pier. Different pier shapes used in the present study are circular, round-nosed, rectangular and sharp-nosed piers. The accuracy and efficiency of the two hybrid models are analyzed and compared with reference to experimental results using model performance indices (MPI) such as correlation coefficient (CC), normalized root-mean-squared error (NRMSE), normalized mean bias (NMB) and Nash–Sutcliffe efficiency (NSE). The ANFIS model with Gbell membership and the PSO–SVM model with polynomial kernel function yield good results in terms of MPI. The performance of PSO–SVM with polynomial kernel function with CC of 0.949, NRMSE of 7.47, NMB of ? 0.009 and NSE of 0.90 reveals that the hybrid ANFIS model with Gbell membership function yields slightly better than that of the PSO–SVM model with CC of 0.950, NRMSE of 6.92, NMB of ? 0.002 and NSE of 0.91 for the optimum bridge pier with circular shape, whereas the performance of PSO–SVM model is better than that of ANFIS model for optimum bridge piers with rectangular and sharp nose shape. The PSO–SVM model can be adopted as accurate and efficient alternative approach in predicting scour depth of the pier. © 2018, The Natural Computing Applications Forum.Item Compressive strength prediction of SCC containing fly ash using SVM and PSO-SVM models(Structural Engineering Research Centre, 2021) Rajeshwari, R.; Mandal, S.; C, C.Self-Compacting Concrete (SCC), is a highly workable material, compacted by its self weight without observable segregation and bleeding. In this study, Support Vector Machine (SVM) and particle swarm optimization based SVM models are employed to predict the 28 days compressive strength of individual SCC mix. A database of 62 no’s of SCC compressive strength from literature with cement partially replaced by fly ash is used for training the models. The test data consists of two groups, an individual study consisting of 9 datasets and other combination of three studies with 19 datasets tested separately. Similar input parameters from the train data is extended for testing the models prediction accuracy. Statistical parameters such as correlation coefficient, root mean square error and scatter index are used to evaluate the models’ prediction results. The particle swarm optimization based SVM model is capable of selecting appropriate SVM parameters to increase the prediction accuracy. From the results, it is seen that both SVM and particle swarm optimized SVM models have good capability in predicting the SCC compressive strength. © 2021, Structural Engineering Research Centre. All rights reserved.