Comparison of TOPSIS and MAUT methods for recipient determination home surgery

Received Mar 5, 2020 Revised Jul 28, 2021 Accepted Aug 10, 2021 House renovation is given by the government to the community, one of which is the assistance provided in the district. Long Mevery especially Tanah Abang Village, namely House Renovation Assistance. So, it is necessary to implement a DSS in determining the recipient of home renovation assistance by comparing multi-atribute utility theory (MAUT) method and TOPSIS to assist the government in determining the right home renovation assistance recipient. There are 16 criteria and their weight values. This study uses the multi-attribute utility theory (MAUT) method and the order of preference technique based on the similarity to the ideal solution (TOPSIS) as a calculation method to produce output and determine the level of accuracy of each method. The test in this study uses a confusion matrix and compares real data testing with the results of calculations on the system. The results of system testing using MAUT and TOPSIS methods, the accuracy of the MAUT method is 94.28% and the TOPSIS method is 35.71%.


INTRODUCTION
Home renovation is a form of housing assistance given to the community that aims to be used for the sake of a comfortable and livable survival. Provision of Home Surgery is carried out selectively in accordance with established criteria. However, the acceptance of this assistance cannot yet be determined objectively so that it is not on target due to the large number of potential recipients as well as the criteria along with the weight in determining decisions. These problems are not in accordance with Law No. 13 of 2011 and Regulation of the Minister of Social Affairs of the Republic of Indonesia Number 20 of 2017 Concerning "Social Rehabilitation of Inadequate Houses and Environmental Infrastructure Facilities" [1]. So that the need for a computer technology-based decision system method in order to be able to calculate the number of criteria caused by the large number of prospective beneficiaries being recorded by the relevant agencies. In this study, researchers used the multi-attribute utility theory (MAUT) method and the technique for order of preference by similarity to ideal solution (TOPSIS) method, where the two methods will be compared to the level of accuracy when implemented in the case of the study with the aim of providing accurate recommendations. MAUT method is an ease in solving various decision-making problems based on attributes is one of the strengths of this method. And this provides an accurate and realistic result [2]. Based on comparative research conducted with the simple additive weighting and TOPSIS methods to support the selection decision for lecturer admissions [3] as well as the comparison of weighted product (WP) and MAUT methods in the support system for labor recruitment decision decisions [4]. The purpose of this study is to make a comparison between the MAUT and TOPSIS methods [5], [6] to find out which is more accurate and efficient and build a decision support system that compares the two methods for the selection of recipients of Home Surgery Assistance. This system is to help determine the decision of recipients of house renovation assistance in Long Mesangat District, specifically Tanah Abang Village, by comparing the MAUT method and the TOPSIS method. Comparison of methods is done to see which method is the best and approaching maximum results in accordance with existing criteria.

RESEARCH METHOD 2.1. The technique for order of preference by similarity to ideal solution (TOPSIS)
TOPSIS uses the principle that the chosen alternative must have the shortest distance from the positive ideal solution and the longest distance (the farthest) [7], from the negative ideal solution from a geometric point of view using the Euclidean distance (distance between two points) to determine the relative proximity of an alternative to the optimal solution [8], [9]. TOPSIS is based on the concept of where the alternative chosen is not only the best alternative because it has the shortest distance from the ideal solution, but also has the longest the distance from the negative ideal solution. The steps of the TOPSIS algorithm are as follows: Determining the ranking of each alternative TOPSIS requires a ranking of the performance of each alternative Ai on each normalized Cj criteria, namely: Create a weighted normalized decision matrix (2): yij= wi.rij (2) with i = 1,2…m and j = 1,2,…..n. Determine the ideal positive and negative solutions. The positive ideal solution A + and the negative ideal solution A-can be determined based on the normalized weight ranking as (3), Calculate distances with the ideal solution. Alternative distances with positive ideal solutions, Alternative distances with positive ideal solutions are calculated using the formula (6); Determine the preference value for each alternative. The preference value for each alternative is given as (7):

Multy atribute utility theory (MAUT)
MAUT is a quantitative comparison method that usually combines measurement of different risk costs and benefits [10]. The MAUT method is used to convert several interests into numerical values on a scale of 0-1 with 0 representing the worst value and 1 the best value [11]. The steps in the MAUT process are: Create a decision framework, by defining the problem, generate (generate) alternatives that might solve the problem, make a list of all aspects that influence the decision, give weight to every aspect that is there [12]. Existing weights must reflect how important these aspects are to the problem, give also the weight of the alternatives. For each alternative, determine how satisfying the alternative is for each aspect and The evaluation process of each alternative on the aspects that exist to get a decision. In the multi-attribute utility theory method, it is used to convert from multiple interests into numerical valueson a scale from 0-1 with 0 representing the worst choice and 1 being the best [13], [14]. This allows a direct comparison of various measures. The overall evaluation value can be defined by (8): Matrix normalization in (9):

Confusion matrix multi-class
A confusion matrix [15]- [17]. That is, after a classifier has been trained, the confusion matrix produced by this classifier on a validation set could be used to find which classes present some confusion in the classification, and then a more specialised classification structure could be generated [18]. There are 4 (four) terms as a representation of the results of the classification process, the four terms are true positive (TP), true negative (TN), false positive (FP) and false negative (FN) [19]. True positive (TP) is the amount of positive data obtained correctly. True negative (TN) value is the amount of negative data collected correctly. The confusion matrix model can be seen in Table 1 [20].

Actual
Classified as

Data collection and development system
System development method in the case study of determining home surgery using a comparison of the MAUT method and TOPSIS using the Linear Sequential model or commonly called the Waterfall model [21]. This waterfall model process can be developed with research cases based on data requirements, design planning, implementation and the results of research in the form of a system to provide the expected results on This research is in the form of accuracy of the deadly method and the method of topsis for the determination of the recipient of home surgical assistance. The Waterfall method is a structured model, in which there are sequential stages of work and cannot repeat or continue if the previous stage has not been completed. The stages to be carried out by the waterfall model method [22], [23].
In Figure 1 shows in appendix, it starts with providing input in the form of 70 prospective recipients of data on home surgery assistance, and determining criteria. In this study, there were 16 criteria which weighed each criterion. Next, each criterion calculation uses each of the two methods MAUT and TOPSIS separately [24], [25]. After doing the calculations, a confussion Matrix test will be performed to obtain the value of accuracy, precision, and recall on each method. After that, an accurate method for this research case study will be obtained.

Data analysis
Data analysis provides information on the criteria as material for selection in the form of numerical values in accordance with the results of interviews with the Head of the Long Mesangat District Government Section and the Head of Community Ability in Tanah Abang Village. In this research method, there are weights and criteria needed to determine the process of prospective recipients of Home Surgical Assistance. From the criteria determined there are weight values of 16 criteria reaching 100% with several rating groupings. Data analysis based on the results of interviews also obtained 70 sample data to be used in testing the system created. Sampling data is taken from the results of the manual recapitulation conducted by the Head of Community Welfare Affairs in Tanah Abang Village, Long Mesangat District in 2017 and 2018. The files obtained contain the identity of the community along with the verification score data that has been filled out by the Committee and Audit Team. In this Table 2 there are 16 criteria and Figure 2 the determined cost and benefit values as well as the appearance of criteria in the application that has been built.  Figure 2. Display criteria and weight on DSS aplications

Research results
This study, has 16 criteria and each criterion has a weight value obtained from interviews with the Head of Community Welfare Affairs in Tanah Abang Village, Long Mesangat District, East Kutai, where 16 criteria are calculated along with 70 prospective recipients of data on house reconstruction assistance to be compared with the accuracy of the method MAUT and the TOPSIS Method. After getting the results of each MAUT and TOPSIS method calculations, proceed to test the level of accuracy using the confusion matrix test and to test the comparison of original data with the number of 70 prospective recipients of house surgery assistance with data that has been calculated using each of the two methods. Confusion matrix test results with each method are in the Tables 3 and 4.
The test results in Tables 3 and 4 using confussion martix show the accuracy value obtained by the MAUT method 92.28% between the value of the system test with the actual value, 97.56% precision of the accuracy of user requests with answers generated by the system, and recall 93, 02% success rate in finding back information. Whereas the TOPSIS method obtained an Accuracy value of 32.85% between the value of the test with the actual value, Precision 46.93% of the determination of the user's request with the answers   The results of comparison of original data with data managed by the system using the MAUT method and the TOPSIS method can be seen the difference in accuracy results that stand out from the two methods. The MAUT method reaches 94.28% while the TOPSIS method only reaches 32.85% of these results. It is known that the MAUT method is more accurate in processing data on the home surgical assistance recipient determination system. There are several factors that affect the accuracy of the TOPSIS method is lower, namely in the TOPSIS method there are grouping types of cost attribute criteria and benefits so that the results of manual calculations with the calculation of the TOPSIS method are inversely proportional. While in the MAUT method there is no grouping of types of criteria so that the results of the manual calculation with the results of the MAUT method are not much different.

CONCLUSION
Based on the results of research on the selection system for the acceptance of home surgical assistance, there are conclusions. Results from comparing the data of results of home surgery recipients with the results of the recommendation from the system is the application of the method MAUT in this case giving an accuracy result of 94.28%. The application of the method TOPSIS in this case gives an accuracy result of 32.85% from 70 total data. Test results using confixion martix shows the accuracy value obtained by the MAUT method is 92.28% as the accuracy value, 97.56% precision value and given the success rate in finding an information of 93.02%. While the TOPSIS method obtained an accuracy value of 32.85% between the test score and the actual value, precision 46.93% of determining user requests with answers generated by the system, and given the 53.48% success rate in rediscovering information. The high accuracy of the method MAUT is due in original data from the Head of Public Welfare Affairs perform the calculation process by adding up the entire score without grouping the type of benefit or cost criteria the same as the MAUT method calculation, so the results of the calculation are not much different. The low accuracy of the TOPSIS method is because in the original data from the Head of Public Welfare Affairs Mr. Hidayatullah there was no grouping of the type of criteria namely benefits and costs so that the calculation results are inversely proportional. So from the result of the test that have been Carried out, The MAUT method is more accurate to provide recommendations on the determination of home surgical recipients. Figure 1. System acceptance of home surgical assistance