目前臺灣營建產業現況係屬於承攬性產業，工地現場會將各工項分別分包予專業協力廠商之比例非常之高，而不同協力廠商之專業程度或是配合程度將是專案成功之重要因素，因此如何根據協力廠商過往表現來選擇適宜之廠商為目前重要課題。但協力廠商各項目之評鑑與其專案最終整體表現之相互對映關係並不明確，如果僅以決策者經驗方式判斷，往往會侷限於個人主觀意見。因此如何改善此一問題，亦為本研究需探討之方向。
有鑑於近年來人工智慧與最佳化模式已成功應於各領域決策輔助課題，本研究將透過收集營建專業協力廠商評鑑分數與歷史表現案例，並導入應用SOS-LSSVM預測模式來歸納出其相互映射之關係，並預測該廠商在新專案之表現。
此外，營造業在面臨新廠商之遴選時，表現能力最好的廠商通常報價不一定是最低，而報價則是營造廠在選商時亦是很重要因素之一，營造廠在預算及利潤之限制下，勢必需要在價格與施工表現中做取捨。因此，本研究將使用SOS最佳化模式在同時考量廠商表現與報價組合之綜合表現下進行廠商組合遴選。最後透過案例驗證，所建立營造廠專案廠商評鑑預測與組合遴選模式之結果與實際採行方式大致相符，部分廠商遴選結果因為考量成本優先之狀況下而有差異，使用者可調整成本或表現偏好作為後續決策之有效參考依據。

The current status of Taiwan’s construction industry is a contracting industry. The percentage of construction sites that will be subcontracted to professional third-party vendors is very high. The professionalism or cooperation of different third-party vendors will be an important factor in the success of the project. Therefore, how to select a suitable subcontractor based on the past performance of the third party is an important issue at present. However, the relationship between the various evaluations of the manufacturer and the final performance of the project is not clear. If it is judged only by the experience of the decision maker, it will often be limited to personal subjective opinions. Therefore, how to solve this problem is also the direction of this research.
In view of the fact that in recent years, artificial intelligence and optimization models have been successfully applied to decision-making support topics in various fields, this study will summarize the evaluation scores and historical performance cases of professional contractors, and introduce and apply the SOS-LSSVM prediction model Mutually map the relationship and predict the performance of the vendor in the new project.
In addition, when the construction industry is faced with the selection of new subcontractors, the quotation of the subcontractors with the best performance is usually not necessarily the lowest, and the quotation is one of the most important factors when the construction plant selects a business. Make a trade-off in construction performance. Therefore, in this study, the SOS optimization model will be used to select the subcontractors portfolio under the consideration of the comprehensive performance of both the vendor performance and the quotation portfolio. Finally, through case verification, the results of the evaluation and prediction and combination selection model of the project of the established construction plant are generally consistent with the actual adoption method. Some subcontractors' selection results are different due to the consideration of cost priority. The user can adjust the cost or performance Preferences serve as an effective reference basis for subsequent decisions.

[1] 李怡軒，「應用層級分析法建構廠商評選因子之研究」，碩士論文，國立高雄科技大學營建工程系研究所，台北，2018。
[2] 陳鴻震，「營造廠專業協力廠商評鑑模式之建立-應用演化式支持向量機推論模式」，碩士論文，國立臺灣科技大學營建工程系研究所，台北，2008。
[3] 黃忠發，「營造業分包管理策略中班底效應之研究」，博士論文，國立中央大學土木工程研究所，桃園，2000。
[4] 吳宗桂，「營造廠專業協力廠商評鑑模式之建立與應用」，碩士論文， 國立台灣科技大學營建工程系，台北，2001。
[5] Min-Yuan Cheng, Doddy Prayogo*, Yu-Wei Wu, “A self-tuning least squares support vector machine for estimating the pavement rutting behavior of asphalt mixtures”, Soft Computing, Vol.23(17), pp.7755-7768, SCI, 2019。
[6] Min-Yuan Cheng, Doddy Prayogo*, “Symbiotic Organisms Search: A new metaheuristic optimization algorithm”, Computers & Structures 139, 98-112, 2014。
[7] 張繼忠，「建築工程分包管理之研究」，碩士論文，國立台灣工業技術學院營建工程技術研究所，台北，1996
[8] 郭慶鴻，「公共工程分包管理制度之研究」，碩士論文，國立台灣科技大學營建工程系，台北，2000。
[9] 營建自動化策略聯盟，「營建協力廠商與材料廠商資訊系統使用手冊」，台北，1999。
[10] G. Edward Gibson, Jr. Michele R. Hamilton. “Analysis of Pre-Project
Planning Effort and Success Variables for Capital Facality Projects ”, ,The
University of Texas at Austin , November 1994.
[11] G. Edward Gibson Jr. Aniello L. Tortora, Charles T. Wilson.“Perceptions
of Project Representatives Concerning Project Success And Pre-Project
Planning Effort“,The University of Texas at Austin , September 1994.
[12] 吳麗娟，「公共工程施工性查核評估模式之建立」，碩士論文，國立台
灣科技大學營建工程系，台北，1998。
[13] 楊惇安，「混凝土強度預測暨最佳化經濟性配比模式之研究」，碩士論
文，國立台灣科技大學營建工程系，台北，2017。
[14] 王俊堯，「應用自動調適生物共生演算法於二維鋼板切割最佳化之研
究」，碩士論文，國立台灣科技大學營建工程系，台北，2018。
[15] MAI THI NGOC DAO，「Artificial Intelligence-Integrated Approach to
Optimize the Mixture Design of High Performance Concrete under
Desired Properties」，National Taiwan University of Science and
Technology，2016。
[16] HANH THI MINH NGUYEN，「Novel Time-Cost Crash Optimization
Algorithm in Penalty-Driven Schedule Delays for Managing Construction
Projects」，National Taiwan University of Science and Technology，2019。
[17] 傅俊凱，「設備保養排程最佳化之探討 -以麵粉生產工廠為例」，碩士
論文，國立台灣科技大學營建工程系，台北，2019。