Faculty Publications
Permanent URI for this communityhttps://idr.nitk.ac.in/handle/123456789/18736
Publications by NITK Faculty
Browse
23 results
Search Results
Item Hierarchical clustering algorithm for land cover mapping using satellite images(2012) Senthilnath, J.; Omkar, S.N.; Mani, V.; Tejovanth, N.; Diwakar, P.G.; Archana Shenoy, B.This paper presents hierarchical clustering algorithms for land cover mapping problem using multi-spectral satellite images. In unsupervised techniques, the automatic generation of number of clusters and its centers for a huge database is not exploited to their full potential. Hence, a hierarchical clustering algorithm that uses splitting and merging techniques is proposed. Initially, the splitting method is used to search for the best possible number of clusters and its centers using Mean Shift Clustering (MSC), Niche Particle Swarm Optimization (NPSO) and Glowworm Swarm Optimization (GSO). Using these clusters and its centers, the merging method is used to group the data points based on a parametric method (k-means algorithm). A performance comparison of the proposed hierarchical clustering algorithms (MSC, NPSO and GSO) is presented using two typical multi-spectral satellite images - Landsat 7 thematic mapper and QuickBird. From the results obtained, we conclude that the proposed GSO based hierarchical clustering algorithm is more accurate and robust. © 2012 IEEE.Item MPI-based parallel synchronous vector evaluated particle swarm optimization for multi-objective design optimization of composite structures(2012) Omkar, S.N.; Venkatesh, A.; Mudigere, M.This paper presents a decentralized/peer-to-peer architecture-based parallel version of the vector evaluated particle swarm optimization (VEPSO) algorithm for multi-objective design optimization of laminated composite plates using message passing interface (MPI). The design optimization of laminated composite plates being a combinatorially explosive constrained non-linear optimization problem (CNOP), with many design variables and a vast solution space, warrants the use of non-parametric and heuristic optimization algorithms like PSO. Optimization requires minimizing both the weight and cost of these composite plates, simultaneously, which renders the problem multi-objective. Hence VEPSO, a multi-objective variant of the PSO algorithm, is used. Despite the use of such a heuristic, the application problem, being computationally intensive, suffers from long execution times due to sequential computation. Hence, a parallel version of the PSO algorithm for the problem has been developed to run on several nodes of an IBM P720 cluster. The proposed parallel algorithm, using MPI's collective communication directives, establishes a peer-to-peer relationship between the constituent parallel processes, deviating from the more common master-slave approach, in achieving reduction of computation time by factor of up to 10. Finally we show the effectiveness of the proposed parallel algorithm by comparing it with a serial implementation of VEPSO and a parallel implementation of the vector evaluated genetic algorithm (VEGA) for the same design problem. © 2012 Elsevier Ltd. All rights reserved.Item Improved network survivability using multi-threshold adaptive range clustering (M-TRAC) algorithm for energy balancing in wireless sensor networks(2013) Shivaprakasha, K.S.; Kulkarni, M.; Joshi, N.Wireless Sensor Networks (WSNs) have become one of the most prominent areas of research in the field of modern communication systems. But unlike IP based routing, WSNs focus on data centric communication. Thus routing is one of the important issues to be considered for a WSN. As sensor networks are generally deployed in hostile environments, batteries cannot be recharged often. Thus energy conservation is one of the important design parameters for WSNs. Many energy aware routing protocols were proposed in the literature. Cluster based algorithms were proved to be better compared to multi-hop routing. In this paper an attempt is made in proposing a novel cluster based energy efficient routing algorithm for WSNs namely the Multi-Threshold Adaptive Range Clustering (M-TRAC) algorithm, which incorporates centralized network management with variable thresholds in order to assure a uniform load distribution amongst nodes, thus improving the network lifetime. © 2013 - IOS Press and the authors. All rights reserved.Item A bio-inspired, incremental clustering algorithm for semantics-based web service discovery(Inderscience Enterprises Ltd., 2015) Kamath S?, S.; Ananthanarayana, V.S.Web service discovery is a challenging task due to the widespread availability of published services on the web. In this paper, a service crawler-based web service discovery framework is proposed, that employs information retrieval techniques to effectively retrieve available, published service descriptions. Their functional semantics is extracted for similarity computation and tag generation using natural language processing techniques. The framework is inherently dynamic in nature as new service descriptions may be continually added during periodic crawler runs or existing ones may be removed if service is unavailable. To deal with these issues, a dynamic, incremental clustering approach based on bird flocking behaviour is proposed. Experimental results show that semantic analysis and automatic tagging captured the services' functional semantics in a meaningful way. The algorithm effectively handled the dynamic requirements of the proposed framework by eliminating cluster recomputation overhead and achieved a speed-up factor of 61.8% when compared to hierarchical clustering. © 2015 Inderscience Enterprises Ltd.Item Prediction model for peninsular Indian summer monsoon rainfall using data mining and statistical approaches(Elsevier Ltd, 2017) Vathsala, H.; Koolagudi, S.G.In this paper we discuss a data mining application for predicting peninsular Indian summer monsoon rainfall, and propose an algorithm that combine data mining and statistical techniques. We select likely predictors based on association rules that have the highest confidence levels. We then cluster the selected predictors to reduce their dimensions and use cluster membership values for classification. We derive the predictors from local conditions in southern India, including mean sea level pressure, wind speed, and maximum and minimum temperatures. The global condition variables include southern oscillation and Indian Ocean dipole conditions. The algorithm predicts rainfall in five categories: Flood, Excess, Normal, Deficit and Drought. We use closed itemset mining, cluster membership calculations and a multilayer perceptron function in the algorithm to predict monsoon rainfall in peninsular India. Using Indian Institute of Tropical Meteorology data, we found the prediction accuracy of our proposed approach to be exceptionally good. © 2016 Elsevier LtdItem Spatial–Temporal Aspects Integrated Probabilistic Intervals Models of Query Generation and Sink Attributes for Energy Efficient WSN(Springer New York LLC barbara.b.bertram@gsk.com, 2017) Kumar, P.; Chaturvedi, A.With advancement in device miniaturization and efficacy of network protocols, in a variety of civilian and military applications, wireless sensor networks (WSNs) architectures find room as viable network paradigm. Invariably, in all these WSN architectures, devising suitable algorithms for the efficient network resources utilization has been a challenging task. In certain events driven scenarios, random arrival pattern of queries generation; their geographical distribution (spatial aspect) and generation rate (temporal aspect) are hard to predict precisely. However, these phenomenons could be appropriately modelled using probabilistic framework while yielding adequate accuracy. Usually, in adopted probabilistic models, the associated control parameters are treated as crisp numbers, which fail to encompass uncertainties that are inevitably associated with the modeled parameters. To include impact of such uncertainties, we propose a modified Poisson PMF expressions in that dependency on spatial and temporal aspects is incorporated based on interval concepts. The paper also validates the dynamic fuzzy c-means algorithm as the most efficient clusters formation scheme. Sink node is an important entity/interface between end users and remotely located sensor nodes. To exploit implications of sink nodes attributes, three different case studies are presented. Wherein, we explore the network surveillance by a single stationary/portable sink and four stationary sinks. Obtained simulation results are analyzed for different scenarios which in principle governed by usage of four distinct clustering schemes and sink(s) attribute driven network surveillance. © 2017, Springer Science+Business Media New York.Item EGA-FMC: Enhanced genetic algorithm-based fuzzy k-modes clustering for categorical data(Inderscience Enterprises Ltd., 2018) Narasimhan, M.; Balasubramanian, B.; Kumar, S.D.; Patil, N.Categorical data clustering is the unsupervised technique of grouping similar objects which have categorical attributes. We propose a genetic algorithm-based fuzzy k-modes categorical data clustering algorithm using multi-objective rank-based selection with enhanced elitism operation. Compactness of the clusters and inter-cluster separation were chosen as objectives to be optimised. During elitism, in every iteration, the best parent chromosomes were identified. The entire population was passed through the selection, crossover and mutation steps. The worst children were then replaced by the best parents. Our method was evaluated on three real-world datasets and resulted in clusters of better quality as compared to current methods with a significant reduction in computation time. Additionally, statistical significance tests were conducted to show the superiority of our approach over other clustering solutions. © © 2018 Inderscience Enterprises Ltd.Item Rough fuzzy joint probabilistic association fortracking multiple targets in the presence of ECM(Elsevier Ltd, 2018) Satapathi, G.S.; Srihari, P.A novel rough fuzzy joint probabilistic data association algorithm (RF-JPDA) is presented to improve the performance of multitarget tracking in the presence of clutter, electronic countermeasures (ECM) and false alarms. The possibility data association matrix is evaluated by applying upper and lower approximations of validated measurements which are obtained from the radar. Four case studies are taken to validate the proposed data association algorithm. The proposed technique performance has been compared with conventional joint probabilistic data association filter (JPDA), fuzzy clustering means (FCM), and fuzzy Genetic Algorithm (Fuzzy-GA) approaches. A hybrid data association approach is formulated and examined for multi-target tracking using intelligent technique. Further, it is evident from the experimental results that RF-JPDA approach is providing enhanced performance in terms of position root mean square error (RMSE), velocity RMSE and execution time for all cases. The average position and velocity RMSE of RF-JPDA are 42.3% and 16.98% less when compared to conventional JPDA. Thus accomplishing novel and effective multiple target tracking algorithm based on expert systems. © 2018 Elsevier LtdItem Hierarchical clustering approaches for flood assessment using multi-sensor satellite images(Taylor and Francis Ltd. michael.wagreich@univie.ac.at, 2019) Senthilnath, J.; Shreyas, P.B.; Rajendra, R.; Sundaram, S.; Kulkarni, S.; Benediktsson, J.A.In this paper, hierarchical clustering methods are used on synthetic aperture radar (SAR) (during the flood) and LISS-III (before the flood) data to analyse damage caused by floods. The flooded and non-flooded regions are extracted from the SAR image while different land cover regions are extracted from the LISS-III image. Initially, the Bayesian information criterion (BIC) is implemented to obtain the constraints for the number of clusters. The optimal cluster centres are then computed using hierarchical clustering approach (i.e. cluster splitting and merging techniques). The cluster splitting techniques such as Iterative Self-Organising Data Technique (ISODATA), Mean Shift Clustering (MSC), Niche Genetic Algorithm (NGA) and Niche Particle Swarm Optimisation (NPSO) were applied on SAR and LISS-III data. The cluster centres obtained from these algorithms are used to group similar data points by using merging method into their respective classes. Further, the results obtained for each method are overlaid to analyse the individual land cover region that is affected by floods. © 2018, © 2018 Informa UK Limited, trading as Taylor & Francis Group.Item A novel filter–wrapper hybrid greedy ensemble approach optimized using the genetic algorithm to reduce the dimensionality of high-dimensional biomedical datasets(Elsevier Ltd, 2019) Gangavarapu, T.; Patil, N.The predictive accuracy of high-dimensional biomedical datasets is often dwindled by many irrelevant and redundant molecular disease diagnosis features. Dimensionality reduction aims at finding a feature subspace that preserves the predictive accuracy while eliminating noise and curtailing the high computational cost of training. The applicability of a particular feature selection technique is heavily reliant on the ability of that technique to match the problem structure and to capture the inherent patterns in the data. In this paper, we propose a novel filter–wrapper hybrid ensemble feature selection approach based on the weighted occurrence frequency and the penalty scheme, to obtain the most discriminative and instructive feature subspace. The proposed approach engenders an optimal feature subspace by greedily combining the feature subspaces obtained from various predetermined base feature selection techniques. Furthermore, the base feature subspaces are penalized based on specific performance dependent penalty parameters. We leverage effective heuristic search strategies including the greedy parameter-wise optimization and the Genetic Algorithm (GA) to optimize the subspace ensembling process. The effectiveness, robustness, and flexibility of the proposed hybrid greedy ensemble approach in comparison with the base feature selection techniques, and prolific filter and state-of-the-art wrapper methods are justified by empirical analysis on three distinct high-dimensional biomedical datasets. Experimental validation revealed that the proposed greedy approach, when optimized using GA, outperformed the selected base feature selection techniques by 4.17%–15.14% in terms of the prediction accuracy. © 2019 Elsevier B.V.
- «
- 1 (current)
- 2
- 3
- »