International Journal of Electrical and Computer Engineering (IJECE) Vol. 7, No. 2, April 2017, pp. 858~868 ISSN: 2088-8708, DOI: 10.11591/ijece.v7i2.pp858-868  858 Journal homepage: http://iaesjournal.com/online/index.php/IJECE The Weights Detection of Multi-Criteria by using Solver Fachrurrazi 1 , Yuwaldi Away 2 , Saiful Husin 3 1,3 Syiah Kuala University Department of Civil Engineering, Indonesia 2 Syiah Kuala University Department of Electrical Engineering, Indonesia Article Info ABSTRACT Article history: Received Dec 13, 2016 Revised Mar 16, 2017 Accepted Mar 30, 2017 Multi criteria, which are generally used for decision analysis, have certain characteristics that relate to the purpose of the decision. Multi criteria have complex structures and have different weights depending upon the consideration of assessors and the purpose of the decision also. Expert’s judgment will be used to detect the criteria weights that applied by assessors. The aim of this study is a model to detect the criteria weights and biases on the subcontractor selection and detecting the significant weights, as decisive criteria. A method, which is used to modeling the weights detection, is the Solver Application. Data, totaling 40 sets, has been collected that consist of the assessor’s assessment and the expert’s judgment. The result is a pattern of weights and biases detection. The proposed model have been able to detect of 20 criteria weights and biases, that consist of 4 criteria in the total weights of 60% (as decisive criteria) and 16 criteria in the total weights of 40%. A model has been built by training process performed by the Solver, which the result for MSE training is 9.73711e-08 and for MSE validation is 0.00900528. Novelty in the study is a model to detect pattern of weights criteria and biases on subcontractor selection by transferring the expert's judgment using Solver Application. Keyword: Bias weights Criteria weights Expert judgment Multi criteria Solver application Subcontracts Copyright © 2017 Institute of Advanced Engineering and Science. All rights reserved. Corresponding Author: Fachrurrazi, Laboratory of Civil Engineering Management, Syiah Kuala University, Indonesia. Email: fachrurrazi@unsyiah.ac.id; fachrurrazi.unsyiah@gmail.com 1. INTRODUCTION The main contractor as a company who is responsible for completing the construction work should be able to act effectively and efficiently. One of the actions to support the result is by partnering with the right subcontractor. Partnering with the subcontractor will provide good result if the partnering is started with the process of qualifying a subcontractor properly, by applying the decision-making procedure correctly [1]. The procedure is important, such as determining the weights and decision criteria. The Assessors, as the persons are doing the evaluation process for the selection of subcontractors, often have differences in determining the criteria weights and sometimes involve subjectivity [1], [2]. The criteria and its weights are not transparent in the selection process [2], sometimes, will make stumped the subcontractor in a strategy to win the bidding proposal. Subcontractor as potential partners must perform the proper analysis for the weights of the criteria that most determine and affect the assessment of assessors. Incomplete data information about the weights of the criteria will cause a problem in analyzing. Based on these background, the problems to be answered in this study are how the weights pattern of the criteria and biases in the decision hierarchy structures of the subcontractor selection that are made by the assessors and how to detect the significant weights as decisive criteria. Implementing of these objectives, it needs the assessment of expert’s judgments that perceived will represent ideal conditions [2]. Various methods and techniques have been conducted to assess the criteria weights, such as Decision Support Systems (DSS) or Expert Systems (ES), generally do not succeed in transferring properly