Kongu Engineering College

About: Kongu Engineering College is a based out in . It is known for research contribution in the topics: Cluster analysis & Control theory. The organization has 2001 authors who have published 1978 publications receiving 16923 citations.

Dynamic Scheduling in Grid Environent with the Improvement of Fault Tolerant Level

Abstract: Objective: The main intent of this research is to improve the fault tolerance capability in the geographically distributed resources by predicting the execution time of every job's earlier in the grid environment. Methods: In this manuscript Fault-Tolerant Based Dynamic scheduling algorithm (FTDS) is introduced for improving the fault tolerance mechanism. The execution time of every user submitted jobs are predicted to improve the fault tolerance capability. The check pointing system is maintained to keep track of all the jobs that are currently running in order to reschedule it in the new resources at the time of resource failure from the last execution step. The resource allocation at the time of failure is decided by the FTDS method and job processing state is decided by the check pointing system. Results: The Fault tolerance based dynamic scheduling algorithm shows better performance than the existing adaptive task check pointing and replication mechanism. In FTDS method, the processing time related parameters are computed to compare it with the existing approach. If the number of task is 500, the average response time in FTDS is 451 ms and the average co-ordination delay is 250ms and the makespan consumed is 450ms. Based on the comparison and the results from the experiment, it proves that the proposed approach works better than the other existing works with better performance. Conclusion: The findings demonstrate that the Fault-Tolerant Based Dynamic Scheduling algorithm is presented and suggested that this method has high fault tolerance rate to prevent the task execution from the resource failure.

Parallel Semi-supervised enhanced fuzzy Co-Clustering (PSEFC) and Rapid Association Rule Mining (RARM) based frequent route mining algorithm for travel sequence recommendation on big social media

Abstract: In this proposed method, with the aim of resolving the frequent route mining issue, by means of utilizing the Rapid Association Rule Mining (RARM) for frequent route mining, recurrently utilized routes as well as smaller distance routes are mined. As a result, support speedy decision of the route identify more willingly than the standard route optimization. Primarily, Multi‐Ontology based Points of Interest (MO‐POIs) model is presented that takes the POIs of user design in combination with the semantic info of the individual users. Furthermore, it as well takes another two steps that are along these lines: (1) routes ranking as per the similarity amid user package as well as routes packages are carried out by means of utilizing Parallel Semi‐supervised enhanced fuzzy Co‐Clustering (PSEFC) as well as (2) route optimizing by Parallel Ant Colony Optimization (PACO) technique in keeping with identical social users' records. The graph model is denoted as the amount of ants in the population of the identical user records. Assess the RARM‐PSEFC recommendation system on a set of Flickr images uploaded by users as well as travel POIs in numerous cities and show its efficiency.

Motor imagery signal classification using semi supervised and unsupervised extreme learning machines

Abstract: Brain computer interface (BCI) aims at providing a brand new communication approach without brain's traditional output through nerve and muscle. “Electroencephalography” has been widely used for BCI system as it is a non-invasive approach. Recently, various classifiers have been used for the analysis of EEG signals measured under the planning and relaxed state. The major work addressed in the paper is the classification of EEG signals (motor imagery) measured under planning and relaxed state using advanced learning classifiers. The dataset of planning and relaxed state is a benchmark data and it is taken from UCI (University of California, Irvine) repository. Semi supervised ELM (SS-ELM) and unsupervised ELM (US-ELM) are recently developed networks and used for the EEG signal classification task. Both of these algorithms can be fit in to a unified framework and handle multi-class classification or multi-cluster clustering. SS-ELM algorithm performs better than US-ELM and other real valued algorithms in classifying planning and relaxed states. The improvement is due to the use of spectral techniques in embedding and clustering.

Hybrid optimal algorithm-based 2D discrete wavelet transform for image compression using fractional KCA

Abstract: Due to the low compression performance of traditional compression models, we have developed a new HOA based Fractional KCA with 2D-DWT for improving the multispectral image quality In this paper, we present a novel multispectral image compression method for improving the complexity by maintaining quality reconstruction and also reducing the size of the storage of multispectral images Initially, Karhunen–Loeve transform (KLT) is used to remove the spatial redundancies In the second stage, 2D DWT is used to eliminate the intraband spatial redundancies In the third stage, Fractional KCA (FKCA) is applied to improve the post-transformation process FKCA is connected to the band of all wavelet sub-bands to minimize the spatial redundancy between intra sub-bands Finally, the Hybrid Optimal algorithm (HOA) based FKCA is used to eliminate the residual and information redundancy among the neighboring bands The experimental analysis of proposed 2D-DWT based Fractional KCA shows that the model improves the performance of compression data in terms of PSNR, MSSI, and VIF Also, the multispectral image dataset shows the proposed compression model outperforms the existing compression models such as FKLT + PCA, ADWT + OADL, and DWT + DCT