Fast Load Balancing Approach for Growing Clusters by Bioinformatics

📝 Abstract

This paper presents Fast load balancing technique inspired by Bioinformatics is a special case to assign a particular patient with a specialist physician cluster at real time. The work is considered soft presentation of the Gaussian mixture model based on the extracted features supplied by patients. Based on the likelihood ratio test, the patient is assigned to a specialist physician cluster. The presented algorithms efficiently handle any size and any numbers of incoming patient requests and rapidly placed them to the specialist physician cluster. Hence it smoothly balances the traffic load of patients even at a hazard situation in the case of natural calamities. The simulation results are presented with variable size of specialist physician clusters that well address the issue for randomly growing patient size.

💡 Analysis

This paper presents Fast load balancing technique inspired by Bioinformatics is a special case to assign a particular patient with a specialist physician cluster at real time. The work is considered soft presentation of the Gaussian mixture model based on the extracted features supplied by patients. Based on the likelihood ratio test, the patient is assigned to a specialist physician cluster. The presented algorithms efficiently handle any size and any numbers of incoming patient requests and rapidly placed them to the specialist physician cluster. Hence it smoothly balances the traffic load of patients even at a hazard situation in the case of natural calamities. The simulation results are presented with variable size of specialist physician clusters that well address the issue for randomly growing patient size.

📄 Content

International conference on Signal Processing, Communication, Power and Embedded System (SCOPES)-2016

Fast Load Balancing Approach for Growing Clusters by Bioinformatics
Soumen Kanrar
Department of Computer Science
Vidyasagar University, Midnapour, WB India rscs_soumen@mail.vidyasagar.ac.in

Abstract—This paper presents Fast load balancing technique inspired by Bioinformatics is a special case to assign a particular patient with a specialist physician cluster at real time. The work is considered soft presentation of the Gaussian mixture model based on the extracted features supplied by patients. Based on the likelihood ratio test, the patient is assigned to a specialist physician cluster. The presented algorithms efficiently handle any size and any numbers of incoming patient requests and rapidly placed them to the specialist physician cluster. Hence it smoothly balances the traffic load of patients even at a hazard situation in the case of natural calamities. The simulation results are presented with variable size of specialist physician clusters that well address the issue for randomly growing patient size.
Keywords—cluster; threshold; feature vector; likelihood; bioinformatics.
I. INTRODUCTION
At the current discard, people need better medical advice at real time. The major issue is to assign the patient to the specified specialist physician without any delay. In the previous work by Nan Liu el. al.[1] have considered the heuristic policies for scheduling patient appointments taking into account the fact that patients may cancel or defer their appointments. Nan Liu el al.[1] have considered various numbers of heuristic policies to present the scheduling model. A. Hertz and N. Lahrichi [2] have addressed the problem in a different way for assigning patients to nurses in the course of home-care services. A. Hertz and N. Lahrichi [2] addressed the workload balance of the nurses, to avoid long travel time for the visit of patients. In this regards, A. Hertz and N. Lahrichi [2] have proposed ‘Tabu’ search algorithm for the patient assignment problem. The ‘Tabu’ search algorithm given a solution space X and a function f that measures the value  f x of every solution x S  , and X S  . Their proposed ‘Tabu’ search algorithm had a specific objective to determine a specific solution *, x which is used to minimize   * f x over X . The obtained minimum value is nothing, but the minimum load assigns to the nurse. The survey paper by Gupta and Denton [3, 4] vigorously focused on the practical issues related to appointment scheduling that provides a review of the state of modeling and optimization. Gupta and Denton [3, 4] addressed to the future directions regarding the necessity of bioinformatics in the area of load balancing. The classification made by Gupta and Denton [3, 4] regarding the research on appointment of scheduling with respect to the type of waiting modeled as direct and indirect. Gupta and Denton[3,4] indicate, most of the existing research has concentrated on direct waiting times. The direct waiting time is the time the patients generally considered to spend waiting in the clinic on that day of appointments. That work typically analyzed to minimize the expected “cost of time” for a day, which is a function of patient’s direct waiting times, and the physician’s idle time or overtime. Scott Levin et. al [5] have founded an important and apparent imbalance in the distribution of load balancing among all physicians working concurrently. Levin and France [6, 7] have considered the work load and communication patterns for individual physicians in emergency working during the periods of high demand. Still the issue is remained challenging one. A new type of approach is required to address the problem in an efficient way. Currently for the speaker identification and verification is done based on bioinformatics. In speaker identification and verification major two types of approach are considered one is Gaussian Mixture model [8, 9, 10]. Another approach is on the base of ‘i-vector’, that is nothing but space and dimension compactness of GMM generated space [11, 12]. Major issue is to handle a certain growth of the patient set. People seek ‘various medical advices’ from the specialist physician. Article [15] addresses the performance of clustering particularly in the mobile domain without considering the patients biological data and information. Particularly in the wireless medium i.e. the patient used to send their biological data and information, softly consider the handoff issue. The parametric estimation for handoff [13] will be considered as the case of those patients. Those are using the smart phone to send their biological data and information. Some cloud base approach can enhance the problem, particularly for private cloud job allocation [14]. The current issue is considered as how

This content is AI-processed based on ArXiv data.