應用人工智慧遷移學習預測營造工程物價指數

Predicting Construction Cost Index using Artificial Intelligence and Transfer Learning

10.6342/NTU202401182

邱民翰

營建物價預測 ； 遷移學習 ； 預訓練與微調 ； 長短期記憶模型 ； 技術分析指標 ； ； Construction cost forecasting ； Transfer Learning ； pre-training and fine-tuning ； LSTM ； Technical Indicator ；

國立臺灣大學土木工程學系學位論文

2024年

碩士

曾惠斌

繁體中文

營建成本超支與工程計價爭議常因材料價格波動而發生，因此需準確預測營建物價，且其可協助預測投標價格、合約管理、估算成本與採購管理等，同時減少工程契約計價爭議。深度學習近年發展快速，但過往較少利用深度學習模型預測營建物價且成果有限，多數研究使用營建物價指數或經濟指標預測短期營建物價。然而，收集經濟指標費時費力，且資料量不足以訓練良好的深度學習模型，而預測短期營建物價與實務較不符合。因此本研究提出使用深度學習結合遷移學習預測營造工程物價指數之框架。利用長短期記憶網絡(Long Short-Term Memory, LSTM)配合預訓練與微調以預測短期與長期之台灣營造工程物價指數，以提升機器學習預測營造工程物價指數之表現。此外，將探討利用單一變數衍生之技術分析指標預測營造工程物價指數之可行性與成效。本研究以營造工程物價總指數、水泥及其製品類指數、金屬製品類指數為目標域，並選擇 7 種匯率作為預選源域，使用 14 種源域選擇方法確定最終源域。模型以 2 層 LSTM 層為基礎，通過預訓練和微調的方法提高預測效能。源域選擇結果顯示，營造工程物價總指數、水泥及其製品類指數之最終源域為新台幣兌日圓之匯率、金屬製品類指數之最終源域為新台幣兌韓元之匯率。預測模型方面，預訓練與微調能夠提升預測營造工程物價指數的效能。在預測營造工程物價總指數與水泥及其製品類指數時，目標域資料集使用技術分析指標比經濟指標好。本研究提出應用預訓練與微調技術預測營造工程物價指數之框架。為首個利用此方法協助預測營造工程物價指數以提高預測表現與泛化能力之研究。這種結合允許模型在新的領域中學習，從而更好地適應不同的營建市場。且此方法優於其他機器學習模型，並能準確預測長期之營造工程物價指數，從而使預測結果更具可靠性和實用性。此外，本研究提出了利用技術分析指標的可行性，並對預測效果進行了探討，進一步豐富了預測方法的多樣性。本研究提出之創新性方法為未來相關研究提供了新的思路，同時也為業界提供了更準確地預測和管理營建成本的手段。這些成果將對營建領域的實踐和研究提供重要參考，有助於提升工程項目的管理效率和成功率。

Construction cost overruns and valuation disputes are common due to fluctuating material prices. Accurate predictions are crucial for forecasting bid prices, contract management, cost estimation, and procurement management. Despite advances in deep learning, its use in predicting construction costs has been limited. Most studies rely on construction cost indices or macroeconomic indicators for short-term forecasts, which are time-consuming and labor-intensive. This study explores the feasibility of technical analysis indicators for more accurate cost predictions and management. It proposes a framework using deep learning and transfer learning, specifically utilizing Long Short-Term Memory (LSTM) networks with pre-training and fine-tuning to predict both short-term and long-term construction cost indices in Taiwan. The research focuses on the general construction cost index, the cement and its products index, and the metal products index as target domains. Seven exchange rates are selected as initial source domains, and 14 source domain selection methods determine the final source domains. The model is based on a two-layer LSTM architecture, with pre-training and fine-tuning enhancing prediction performance. Results indicate that the final source domain for the general construction cost index and the cement and its products index is the New Taiwan Dollar to Japanese Yen exchange rate. In contrast, for the metal products index, it is the New Taiwan Dollar to Korean Won exchange rate. Pre-training and fine-tuning significantly improve prediction performance. Technical analysis indicators outperform macroeconomic indicators for the general construction cost index and the cement and its products index. This study presents a novel framework for predicting construction costs through pre-training and fine-tuning, marking the first application of this method in the field. This approach allows the model to better adapt to different markets, outperforming other machine learning models and providing more reliable long-term predictions. The study showcases technical analysis indicators in construction cost prediction, offering a more accurate method for cost management and project success.

1. [1] 行政院主計總處. 營造工程物價指數, 2024.
2. [2] Akshay Tondak. Recurrent neural networks (rnn) tutorial: Rnn training, advantages disadvantages (complete guidance), 2023. June 23, 2023.
3. [3] Oinkina and Hakyll. Understanding lstm networks, 2015. August 27, 2015.
4. [4] S. J. Pan and Q. Yang. A survey on transfer learning. IEEE Transactions on Knowledge and Data Engineering, 22(10):1345–1359, 2010.
5. [5] L. Torrey and J. Shavlik. Transfer learning, pages 242–264. IGI Global, 2009.
6. [6] 行政院主計總處. 編製方法說明營造工程物價指數. Report, 行政院主計總處, 2023.
7. [7] 林秀貞. 國際油價波動對重要營建材料成本影響之研究 -以鋼筋、水泥、砂石、瀝青為例. Thesis, 國立中央大學, 2007.
8. [8] 謝政達. 營建成本居高 營建技術新思維. 營建知訊, 430:30–36, 2018.
9. [9] S. Hwang, M. Park, H. S. Lee, and H. Kim. Automated time-series cost forecasting system for construction materials. Journal of Construction Engineering and Management, 138(11):1259–1269, 2012.
10. [10] Bent Flyvbjerg, Nils Bruzelius, and Werner Rothengatter. Megaprojects and risk: An anatomy of ambition. Cambridge university press, 2003.
11. [11] Abdulelah Aljohani, Dominic Ahiaga-Dagbui, and David Moore. Construction projects cost overrun: What does the literature tell us? International Journal of Innovation, Management and Technology, 8(2):137, 2017.
12. [12] J. W. Xu and S. Moon. Stochastic forecast of construction cost index using a cointegrated vector autoregression model. Journal of Management in Engineering, 29(1):10–18, 2013.
13. [13] M. A. Musarat, W. S. Alaloul, M. S. Liew, A. Maqsoom, and A. H. Qureshi. Inves- tigating the impact of inflation on building materials prices in construction industry. Journal of Building Engineering, 32:14, 2020.
14. [14] A Uchechukwu Elinwa and Silas A Buba. Construction cost factors in nigeria. Journal of construction engineering and management, 119(4):698–713, 1993.
15. [15] 行政院公共工程委員會. 申訴統計資料 111. Report, 行政院公共工程委員會, 2023.
16. [16] 羅韋淵. 公共工程契約中物價調整機制之問題研究. Thesis, 國立政治大學, 2009.
17. [17] 陳誌泓, 陳信至, and 廖育良. 工程契約中情事變更原則爭議之若干檢討-以 臺灣物價調整款爭議為中心(上). 萬國法律, 249:110–113, 2023.
18. [18] 陳誌泓, 陳信至, and 廖育良. 工程契約中情事變更原則爭議之若干檢討-以 臺灣物價調整款爭議為中心(下). 萬國法律, 250:94–97, 2023.
19. [19] 黃泰鋒 and 陳麗嘉. 營建物價調整常見爭議問題探討. 工程仲裁, 2008.
20. [20] 劉嘉怡. 從相關判決看物價調整款爭議. 營建知訊, 401:45–49, 2016.
21. [21] S.A.M.FaghihandH.Kashani.Forecasting construction material prices using vector error correction model. Journal of Construction Engineering and Management, 144(8):12, 2018.
22. [22] M. Hiransha, E. A. Gopalakrishnan, V. K. Menon, and K. P. Soman. Nse stock market prediction using deep-learning models. In S. Singh, V. K. Asari, R. B. Patel, and P. Sidike, editors, Procedia Computer Science, volume 132, pages 1351–1362. Elsevier B.V., 2018.
23. [23] M. Marzouk and A. Amin. Predicting construction materials prices using fuzzy logic and neural networks. Journal of Construction Engineering and Management, 139(9):1190–1198, 2013.
24. [24] Pushpendu Ghosh, Ariel Neufeld, and Jajati Keshari Sahoo. Forecasting directional movements of stock prices for intraday trading using lstm and random forests. Finance Research Letters, 46:102280, 2022.
25. [25] Luckyson Khaidem, Snehanshu Saha, and Sudeepa Roy Dey. Predicting the direction of stock market prices using random forest. arXiv preprint arXiv:1605.00003, 2016.
26. [26] Sidra Mehtab and Jaydip Sen. Stock price prediction using convolutional neural networks on a multivariate timeseries. arXiv preprint arXiv:2001.09769, 2020.
27. [27] 林楷竣. 以自適應性風險交易策略建構集成式學習投資組合決策系統. Thesis, 國立臺灣科技大學, 2023.
28. [28] PHAMTRANBAOQUYE.Multiobjective-optimized construction stock portfolio investment strategy based on profitability prediction. Thesis, 國立臺灣科技大學, 2023.
29. [29] Yue-Gang Song, Yu-Long Zhou, and Ren-Jie Han. Neural networks for stock price prediction. arXiv preprint arXiv:1805.11317, 2018.
30. [30] Sidra Mehtab, Jaydip Sen, and Abhishek Dutta. Stock price prediction using machine learning and lstm-based deep learning models. In Machine Learning and Metaheuristics Algorithms, and Applications: Second Symposium, SoMMA 2020, Chennai, India, October 14–17, 2020, Revised Selected Papers 2, pages 88–106. Springer, 2021.
31. [31] Ehsan Hoseinzade and Saman Haratizadeh. Cnnpred: Cnn-based stock market prediction using a diverse set of variables. Expert Systems with Applications, 129:273– 285, 2019.
32. [32] A. Maratkhan, I. Ilyassov, M. Aitzhanov, M. F. Demirci, and A. M. Ozbayoglu. Deep learning-based investment strategy: technical indicator clustering and residual blocks. Soft Computing, 25(7):5151–5161, 2021.
33. [33] H Lin, C Chen, G Huang, and A Jafari. Stock price prediction using generative adversarial networks. J. Comp. Sci, pages 17,188–196, 2021.
34. [34] Ching-HwangWangandYong-HoMei.Model for forecasting construction cost indices in taiwan. Construction Management and Economics, 16(2):147 157, 1998. doi: 10.1080/014461998372457.
35. [35] Chi-Young Choi, Kyeong Rok Ryu, and Mohsen Shahandashti. Predicting city-level construction cost index using linear forecasting models. Journal of Construction Engineering and Management, 147(2):04020158, 2021.
36. [36] U.Isikdag, A.Hepsag, S.I.Biyikli, D.Oez, G.Bekdas, and Z.W.Geem. Estimating construction material indices with arima and optimized narnets. Cmc Computers Materials Continua, 74(1):113–129, 2023.
37. [37] 邱敏鋒. 運用支撐向量機建構營建材料供應商使用衍生性金融商品避險之預測模型. 2008.
38. [38] T. Moon and D. H. Shin. Forecasting model of construction cost index based on vecm with search query. Ksce Journal of Civil Engineering, 22(8):2726–2734, 2018.
39. [39] Trefor P. Williams. Predicting changes in construction cost indexes using neural networks. Journal of Construction Engineering and Management, 120(2):306–320, 1994.
40. [40] A. Shiha, E. M. Dorra, and K. Nassar. Neural networks model for prediction of construction material prices in egypt using macroeconomic indicators. Journal of Construction Engineering and Management, 146(3):16, 2020.
41. [41] M. Mir, H. M. D. Kabir, F. Nasirzadeh, and A. Khosravi. Neural network-based interval forecasting of construction material prices. Journal of Building Engineering, 39:13, 2021.
42. [42] 馮重偉 and 江怡萱. 結合深度學習及關鍵字搜尋熱度趨勢於臺灣鋼筋價格漲 跌幅之預測. 中國土木水利工程學刊, 33(8):595–604, 2021.
43. [43] Min-YuanCheng,Nhat-DucHoang,andYu-WeiWu.Hybrid intelligence approach based on ls-svm and differential evolution for construction cost index estimation: A taiwan case study. Automation in Construction, 35:306–313, 2013.
44. [44] Minh-TuCao, Min-YuanCheng, and Yu-WeiWu. Hybrid computational model for forecasting taiwan construction cost index. Journal of Construction Engineering and Management, 141(4):04014089, 2015.
45. [45] B.Q.Tang, J.Han, G.F.Guo, Y.Chen, and S.Zhang.Building material prices forecasting based on least square support vector machine and improved particle swarm optimization. Architectural Engineering and Design Management, 15(3):196–212, 2019.
46. [46] Y. Du, J. Wang, W. Feng, S. Pan, T. Qin, R. Xu, and C. Wang. Adarnn: Adaptive learning and forecasting of time series. In International Conference on Information and Knowledge Management, Proceedings, pages 402–411. Association for Computing Machinery, 2021.
47. [47] Ngoc-Quang Nguyen. Short-term Prediction of Regional Energy Consumption by Jellyfish Search-Optimized Deep Learning Models. Thesis, National Taiwan University of Science and Technology, 2023.
48. [48] M.Ribeiro, K.Grolinger, H.F.ElYamany, W.A.Higashino, and M.A.M.Capretz. Transfer learning with seasonal and trend adjustment for cross-building energy forecasting. Energy and Buildings, 165:352–363, 2018.
49. [49] Haixiang Zang, Lilin Cheng, Tao Ding, Kwok W. Cheung, Zhinong Wei, and Guoqiang Sun. Day-ahead photovoltaic power forecasting approach based on deep convolutional neural networks and meta learning. International Journal of Electrical Power Energy Systems, 118:105790, 2020.
50. [50] S.Raghu, N.Sriraam, Y.Temel, S.V.Rao, and P.L.Kubben. Eegbasedmulti-class seizure type classification using convolutional neural network and transfer learning. Neural Networks, 124:202–212, 2020.
51. [51] N. Strodthoff, P. Wagner, T. Schaeffter, and W. Samek. Deep learning for ecg analysis: Benchmarks and insights from ptb-xl. IEEE Journal of Biomedical and Health Informatics, 25(5):1519–1528, 2021.
52. [52] Hyeong Kyu Choi. Stock price correlation coefficient prediction with arima-lstm hybrid model. arXiv preprint arXiv:1808.01560, 2018.
53. [53] 龔千芬 and 郝沛毅. 融合深度神經網路與深層模糊孿生支持向量機於股價預 測. 資訊管理學報, 29(4):303–333, 2022.
54. [54] W. Bao, J. Yue, and Y. Rao. A deep learning framework for financial time series using stacked autoencoders and long-short term memory. PLoS ONE, 12(7), 2017.
55. [55] MehtabhornObthong,NongnuchTantisantiwong,WatthanasakJeamwatthanachai, and Gary Wills. A survey on machine learning for stock price prediction: algorithms and techniques, 2020.
56. [56] Stéphane Goutte, Hoang-Viet Le, Fei Liu, and Hans-Jörg von Mettenheim. Deep learning and technical analysis in cryptocurrency market. Finance Research Letters, 54:103809, 2023.
57. [57] Sidra Mehtab and Jaydip Sen. A robust predictive model for stock price prediction using deep learning and natural language processing. arXiv preprint arXiv:1912.07700, 2019.
58. [58] MahlaNikou, Gholamreza Mansourfar, and Jamshid Bagherzadeh. Stock price prediction using deep learning algorithm and its comparison with machine learning algorithms. Intelligent Systems in Accounting, Finance and Management, 26(4):164– 174, 2019.
59. [59] Marc Velay and Fabrice Daniel. Stock Chart Pattern recognition with Deep Learning. 2018.
60. [60] Sepp Hochreiter and Jürgen Schmidhuber. Long short-term memory. Neural computation, 9(8):1735–1780, 1997.
61. [61] S. Ben-David, J. Blitzer, K. Crammer, A. Kulesza, F. Pereira, and J. W. Vaughan. A theory of learning from different domains. Machine Learning, 79(1-2):151–175, 2010.
62. [62] J. Lu, V. Behbood, P. Hao, H. Zuo, S. Xue, and G. Zhang. Transfer learning using computational intelligence: A survey. Knowledge-Based Systems, 80:14–23, 2015.
63. [63] K. Weiss, T. M. Khoshgoftaar, and D. D. Wang. A survey of transfer learning. Journal of Big Data, 3(1), 2016.
64. [64] C. Tan, F. Sun, T. Kong, W. Zhang, C. Yang, and C. Liu. A survey on deep transfer learning. In Y. Manolopoulos, B. Hammer, V. Kurkova, L. Iliadis, and I. Magloiannis, editors, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), volume 11141 LNCS, pages 270–279. Springer Verlag, 2018.
65. [65] Jason Yosinski, Jeff Clune, Yoshua Bengio, and Hod Lipson. How transferable are features in deep neural networks? Advances in neural information processing systems, 27:pages 3320–3328, 2014.
66. [66] H. Xu, B. Xu, J. He, and J. Bi. Deep transfer learning based on lstm model in stock price forecasting. In ACM International Conference Proceeding Series, pages 73– 80. Association for Computing Machinery, 2021.
67. [67] Hidetoshi Shimodaira. Improving predictive inference under covariate shift by weighting the log-likelihood function. Journal of statistical planning and inference, 90(2):227–244, 2000.
68. [68] F. Zhuang, Z. Qi, K. Duan, D. Xi, Y. Zhu, H. Zhu, H. Xiong, and Q. He. A comprehensive survey on transfer learning. Proceedings of the IEEE, 109(1):43 76, 2021.
69. [69] E.Otović, M.Njirjak, D.Jozinović, G.Mauša, A.Michelini, and I.S̆tajduhar. Intra-domain and cross-domain transfer learning for time series data—how transferable are the features? Knowledge-Based Systems, 239, 2022.
70. [70] J. Wang, Q. Gu, J. Wu, G. Liu, and Z. Xiong. Traffic speed prediction and congestion source exploration: A deep learning method. In F. Bonchi, J. Domingo-Ferrer, R. Baeza-Yates, Z. H. Zhou, and X. Wu, editors, Proceedings - IEEE International Conference on Data Mining, ICDM, volume 0, pages 499–508. Institute of Electrical and Electronics Engineers Inc., 2016.
71. [71] S. Shao, S. McAleer, R. Yan, and P. Baldi. Highly accurate machine fault diagnosis using deep transfer learning. IEEE Transactions on Industrial Informatics, 15(4):2446–2455, 2019.
72. [72] Z.Cen and J.Wang. Crude oil price prediction model with long short term memory deep learning based on prior knowledge data transfer. Energy, 169:160–171, 2019.
73. [73] J. Ma, J. C. P. Cheng, C. Lin, Y. Tan, and J. Zhang. Improving air quality prediction accuracy at larger temporal resolutions using deep learning and transfer learning techniques. Atmospheric Environment, 214, 2019.
74. [74] M. Kraus and S. Feuerriegel. Decision support from financial disclosures with deep neural networks and transfer learning. Decision Support Systems, 104:38–48, 2017.
75. [75] K. Mishev, A. Gjorgjevikj, I. Vodenska, L. Chitkushev, W. Souma, and D. Trajanov. Forecasting corporate revenue by using deep-learning methodologies. In Proceedings - 2019 3rd International Conference on Control, Artificial Intelligence, Robotics and Optimization, ICCAIRO 2019, pages 115–120. Institute of Electrical and Electronics Engineers Inc., 2019.
76. [76] Q. Q. He, P. C. I. Pang, and Y. W. Si. Multi-source transfer learning with ensemble for financial time series forecasting. In J. He, H. Purohit, G. Huang, X. Gao, and K. Deng, editors, Proceedings - 2020 IEEE/WIC/ACM International Joint Conference on Web Intelligence and Intelligent Agent Technology, WI-IAT 2020, pages 227–233. Institute of Electrical and Electronics Engineers Inc., 2020.
77. [77] T. T. Nguyen and S. Yoon. A novel approach to short-term stock price movement prediction using transfer learning. Applied Sciences (Switzerland), 9(22), 2019.
78. [78] Y. Li, H. N. Dai, and Z. Zheng. Selective transfer learning with adversarial training for stock movement prediction. Connection Science, 34(1):492–510, 2022.
79. [79] 王晋东. 迁移学习导论. 电子工业出版社, 2021.
80. [80] H. I. Fawaz, G. Forestier, J. Weber, L. Idoumghar, and P. A. Muller. Transfer learning for time series classification. In N. Abe, H. Liu, C. Pu, X. Hu, N. Ahmed, M. Qiao, Y. Song, D. Kossmann, B. Liu, K. Lee, J. Tang, J. He, and J. Saltz, editors, Proceedings - 2018 IEEE International Conference on Big Data, Big Data 2018, pages 1367–1376. Institute of Electrical and Electronics Engineers Inc., 2018.
81. [81] 台灣財政部關務署. 台灣財政部關務署海關進出口統計網, 2024. May 01, 2024.
82. [82] 財政部. 鋼鐵業原物料耗用通常水準. Report, 財政部, 2023.
83. [83] 台灣區水泥工業同業公會. 2023 年度台灣區水泥工業同業公會年報. Report, 台灣區水泥工業同業公會, 2023.
84. [84] 張清榮 and 游珊蓉. 營造工程物價指數與國內外大宗物資與金融指數相關性 研究, 2014.
85. [85] 游珊蓉. 營造工程物價指數預測之研究-以徑路分析模式. Thesis, 中華大學, 2015.
86. [86] Mehar Vijh, Deeksha Chandola, Vinay Anand Tikkiwal, and Arun Kumar. Stock closing price prediction using machine learning techniques. Procedia Computer Science, 167:599–606, 2020.
87. [87] J. Patel, S. Shah, P. Thakkar, and K. Kotecha. Predicting stock and stock price index movement using trend deterministic data preparation and machine learning techniques. Expert Systems with Applications, 42(1):259–268, 2015.
88. [88] HS Hota, Richa Handa, and Akhilesh Kumar Shrivas. Time series data prediction using sliding window based rbf neural network. International Journal of Computational Intelligence Research, 13(5):1145–1156, 2017.
89. [89] Shan Zhong and David B Hitchcock. Sp 500 stock price prediction using technical, fundamental and text data. arXiv preprint arXiv:2108.10826, 2021.
90. [90] Baabak Ashuri, Seyed Mohsen Shahandashti, and Jian Lu. Empirical tests for identifying leading indicators of enr construction cost index. Construction Management and Economics, 30(11):917–927, 2012. doi: 10.1080/01446193.2012.728709.
91. [91] 李呈芳, 王小龍, and 黃依典. 營建工程工料變動對營建產業影響之探討. 萬能學報, 31:193–211, 2009.
92. [92] Abimbola Windapo and Keith Cattell. Examining the trends in building material prices: built environment stakeholders' perspectives. Manage Construct Res Pract, 1:187–201, 2012.
93. [93] Vladimir Braverman, Rafail Ostrovsky, and Carlo Zaniolo. Optimal sampling from sliding windows, 2009.
94. [94] JinYang, Hugues Rivard, and Radu Zmeureanu. On-line building energy prediction using adaptive artificial neural networks. Energy and Buildings, 37(12):1250–1259, 2005.
95. [95] Ralf-Peter Mundani, Jérôme Frisch, Vasco Varduhn, and Ernst Rank. A sliding window technique for interactive high-performance computing scenarios. Advances in Engineering Software, 84:21–30, 2015.
96. [96] Jui-Sheng Chou and Thi Truong. Sliding-window metaheuristic optimization-based forecast system for foreign exchange analysis. Soft Computing, 23:3545–3561, 2019.
97. [97] Akash Deep. A multifactor analysis model for stock market prediction. International Journal of Computer Science and Telecommunications, 14(1), 2023.
98. [98] Nitish Srivastava, Geoffrey Hinton, Alex Krizhevsky, Ilya Sutskever, and Ruslan Salakhutdinov. Dropout: A simple way to prevent neural networks from overfitting. Journal of Machine Learning Research, 15(56):1929–1958, 2014.
99. [99] Yitong Duan, Lei Wang, Qizhong Zhang, and Jian Li. Factorvae: A probabilistic dynamic factor model based on variational autoencoder for predicting cross-sectional stock returns. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 36, pages 4468–4476, 2022.
100. [100] 行政院公共工程委員會. 工程採購契約範本(112.11.15 修正). Government document, 行政院公共工程委員會, 2023.