A Method Of Test Cases Based On The Optimization Of DBSCAN Lower Demand Reduction

With the rapid development of software industry,all walks of life have been inseparable from the software support,the quality of the software caused by the loss is also growing.Software testing,as a necessary means of ensuring the quality of software,plays an increasingly important role in the software life cycle.How to carry out efficient software testing has also become the focus of the next study.Software test case reduction as an effective way to improve the efficiency of software testing and reduce the cost of software testing Research on this area is also increasingly important,and various studies seek to achieve certain results in the field of efficient and low-cost testing.This paper systematically reviews various current test case reduction techniques and conducts in-depth research on the basis of test case reduction techniques based on test requirement reduction.A demand reduction based DBSCAN algorithm is proposed Test Case Reduction Technology.Specific research contributions can be broadly summarized as the following aspects:(1)Based on the redundancy of test case set derived from requirement reduction,this paper presents a method to further eliminate redundancy and streamline test cases.The test case reduction method under requirement reduction reduces the size of the test case macroscopically but only modularizes the test case set as a whole for a single requirement rather than the individual test cases within the set between.Therefore,the set of test cases obtained by this method is the result of modular reduction,and there is still room for redundancy elimination.In this paper,a test case reduction method based on the improved DBSCAN algorithm is proposed in this paper,to a certain extent,Redundancy between test cases,played a very good optimization.(2)DBSCAN algorithm as an effective clustering method,the parameter value is the key to improve the algorithm.In this paper,two important parameters of the field of radius E and domain density threshold MinPts value method to improve the effective method proposed based on the distance standard deviation of the domain radius method and based on the minimum spanning tree algorithm domain density threshold The value of the method,greatly improve the scientific and reasonable value of the algorithm to improve the efficiency of the implementation of the algorithm.Based on the distance-standard-deviation domain radius science,which covers the relatively concentrated distribution of data,classifying the data with high similarity and coincidence into one class provides the basis for the next reduction optimization.However,the method based on small spanning tree algorithm for domain density threshold value aggregates the size of the cluster with minimum path cost,which can well cover the data points in high density areas and achieve the appropriate size and coverage of the class.(3)The main function of the traditional DBSCAN clustering algorithm is to output the class,but in the implementation of the algorithm will produce specific core data objects and noise data objects.This paper improves on the traditional algorithm and improves the class output of clustering to the output of data set containing core data objects and noise objects.The specific class in this paper is only used as an optimization precondition and the basic conditions and not optimization object,so there is no need to output all the classes,in order to reduce the time and space complexity of the algorithm and improve the algorithm to solve In this paper,the ability of the actual problem,this paper has been improved on the original algorithm to improve its applicability to the problem,played a role in optimizing the collection of test cases.(4)The specific idea of use case optimization in this paper is based on the idea of equivalence class partitioning and boundary value analysis in black box testing.As the basic theory source of the reduction optimization of test cases,this paper presents two similar reduction strategies,that is,the class equivalence class and the class boundary value analysis strategy.Taking this strategy as the theoretical basis of the optimization of this article,this paper further perfects the theoretical and logical basis of this paper.