A modified particle swarm optimization algorithm for location problem

Abstract

In the Weber location problem which was proposed for optimal location of industrial enterprises, the aim is to find the location of a point such that the sum of weighted distance between this point and a finite number of existing points is minimized. This popular model is widely used for optimal location of equipment and in many sophisticated models of cluster analysis such as detecting homogeneous production batches made from a single production batch of raw materials. The well-known iterative Weiszfeld does not converge efficiently to the optimal solution when the solution either coincides or nearly coincides with one of the demands point which is not the optimum. We propose a modified Particle Swarm Optimization (PSO) algorithm. The velocity update of the PSO is modified to enlarge the search space and enhance the global search ability. The preliminary results of these algorithms are analyzed and compared.

Multi-Objective Modified Particle Swarm Optimization for Test Suite Reduction (MOMPSO)

“…2 [4,21] and the weight factor values from Tab. 1, we calculated the cost involved in processing the whole source code by every test case with the following equation [15,16].…”

“…1 to 6, proposed approach decreases no of test cases in test suite and increases coverage, cost, fault rate and overall efficiency of testing process. By considering the result of different application, proposed approach is more efficient for solving test suite reduction problem [4,15,16,32,34].…”

“…In terms comparing basic PSO, GWO, ACO, Genetic and MOMPSO, the results show that MOMPSO significantly better than other approach because reduction rate, coverage, cost, history, fault and execution time more effective parameter for test case as well as test suite reduction problem. To increasing the performance of algorithm [15][16][17]21,25]. MOMPSO uses multiple parameter such as weight, cost of test case, coverage, fitness of test cases and two search space for finding optimal test cases .Due this reason , parameters which we have considered for reduction algorithm and performance analysis are more efficient for solving test reduction problem [3,17].…”

“…With the selected test cases, the testing team will spend less time for identifying the problem. In this research we focused MOPSO Algorithm for test suite reduction in which we have used effective parameters for selecting the test cases in order to improve cost of selection process [15][16][17][18] .When any meta-heuristic related model is used for test case reduction technique then two very prime principles which must be used in contradiction whenever needed where first principle says that exploration of the space of search is very important and another principle emphasizes on exploiting the optimal solutions [3,6,7].…”

“…Another important parameter which plays a very significant role in regression testing is statement coverage which ensures that the majority of the paths of the given test case execution is performed based on traditional approach which involves some more factors like coverage information including statement [19][20][21], DU pair [22], branch etc. [23][24][25].But still some problem persists in the algorithm like for large systems the coverage of path is a tedious task as it is more time-consuming with respect to parameter of statement coverage or coverage information within the source code [13,16]. The main problem of the usage of this algorithm is its complexity in terms of computation which can be achieved when cyclomatic complexity is less than 10.…”