Loading...
|
Please use this identifier to cite or link to this item:
https://nccur.lib.nccu.edu.tw/handle/140.119/154224
|
| Title: | 運用大型語言模型的提示工程在電子產品的需求預測
Prompt Engineering in Large Language Models for Demand Forecasting of Electronics |
| Authors: | 黃郁清
Huang, Yu-Ching |
| Contributors: | 沈錳坤
Shan, Man-Kwan
黃郁清
Huang, Yu-Ching |
| Keywords: | 大型語言模型
提示工程
電子產品
需求預測
Large Language Models
Prompt Engineering
Electronics
Demand Forecasting |
| Date: | 2024 |
| Issue Date: | 2024-11-01 11:33:09 (UTC+8) |
| Abstract: | 隨著全球經濟起伏和市場變化，供應鏈管理面臨越來越多的挑戰。特別是在電子產品領域，其需求受到生命週期較短、季節性波動及全球化供應鏈等多重因素的影響。和其他生命週期長或是產品生態較穩定的產業相比，這些特性增加了需求預測的複雜性。
電子產品的需求量會受到市場動態頻繁變化加上難以預測的外部因素影響，需要透過多種數據來源 (如分析報告、社交媒體評論或新聞報導等) 綜合評估需求趨勢並精準的分析。傳統基於時間序列分析結構化數據的需求預測方法難以有效應對。為了克服這些挑戰，本文提出使用大型語言模型來提升需求預測的精確性。大型語言模型能夠處理不同來源的非結構化數據，諸如市場趨勢、消費者需求改變或是疫情等事件，這些模型通過自然語言處理技術有效提取和分析關鍵信息，為企業提供更靈活的需求預測模型。
本研究將探討如何通過提示工程將外部動態因子整合至需求預測模型中。我們從多方資料來源蒐集會影響電子產品未來一年每個季度需求量的因子，接著通過特徵選擇選取出對需求預測影響較大的因子組合，並結合提示工程技術應用於需求預測中。實驗結果顯示，運用大型語言模型結合提示工程的技巧，相較於傳統的時間序列模型，能提高電子產品需求預測的準確性。
Supply chain management faces increasing challenges due to economic fluctuations and rapid market changes. The electronics industry, in particular, is significantly impacted by global supply chains, short product life cycles, and seasonal variations. These factors make demand forecasting in this industry more complex compared to ones with longer product life cycles and stable ecosystems.
Electronics demand is influenced by dynamic market forces and unpredictable external factors, necessitating a multifaceted approach to forecasting. Relying solely on traditional demand forecasting methods, which primarily focus on structured time series data, often proves insufficient in capturing these complexities. To overcome these limitations, this paper proposes the use of large
language models (LLMs) to enhance forecasting accuracy by integrating unstructured data sources such as market analyses, social media trends, and global events. LLMs, with their ability to process unstructured data, offer a robust solution for incorporating external dynamic factors, such as shifts in consumer preferences or global disruptions like pandemics. By employing natural language processing techniques, LLMs can extract and analyze valuable insights from these diverse data sources, providing businesses with more flexible and accurate forecasting models.
This study investigates how LLMs can improve demand forecasting by incorporating unstructured external factors into predictive models. We identify key factors from various data sources that impact quarterly demand for electronics, apply feature selection to determine the most significant factors, and utilize prompt engineering techniques to refine forecasting accuracy. Experimental results demonstrate that the integration of LLMs with these advanced techniques
significantly enhances demand forecasting performance in the electronics industry compared to traditional time series models. |
| Reference: | [1] M. Seyedan and F. Mafakheri, Predictive Big Data Analytics for Supply Chain Demand Forecasting: Methods, Applications, and Research Opportunities, Journal of Big Data, 7, 2020.
[2] J. Ma, M. Kwak and H. Kim, Demand Trend Mining for Predictive Life Cycle Design, Journal of Cleaner Production, 68, 2014.
[3] R. Blackburn, K. Lurz, B. Priese, et al., A Predictive Analytics Approach for Demand Forecasting in the Process Industry, International Transactions in Operational Research, 22(3), 2014.
[4] G. Merkuryeva, A. Valberga and A. Smirnov, Demand Forecasting in Pharmaceutical Supply Chains: A Case Study, Procedia Computer Science, 149 (C), 2019.
[5] K. I. Nikolopoulos, M. Z. Babai and K. Bozos, Forecasting Supply Chain Sporadic Demand With Nearest Neighbor Approaches, International Journal of Production Economics, 177(1), 2016.
[6] K. K. Chandriah and R. V. Naraganahalli, Rnn/LSTM with Modified Adam Optimizer in Deep Learning Approach for Automobile Spare Parts Demand Forecasting, Multimedia Tools and Applications, 80(17), 2021.
[7] J. Huber, A. Gossmann and H. Stuckenschmidt, Cluster-Based Hierarchical Demand Forecasting for Perishable Goods, Expert Systems with Applications, 76, 2017.
[8] C. P. Da Veiga, C. R. P. Da Veiga, A. Catapan, et al., Demand Forecasting in Food Retail: A Comparison Between the Holt-Winters and Arima Models, WSEAS Transactions on Business and Economics, 11(1), 2014.
[9] C. L. Yang and H. Sutrisno, Short-Term Sales Forecast of Perishable Goods for Franchise Business, in 10th International Conference on Knowledge and Smart Technology (KST), 2018.
[10] P. Ramos, N. Santos and R. Rebelo, Performance of State Space and Arima Models for Consumer Retail Sales Forecasting, Robotics and Computer-Integrated Manufacturing, 34(17), 2015.
[11] S. P. Jun, D. H. Park and J. Yeom, The Possibility of Using Search Traffic Information to Explore Consumer Product Attitudes and Forecast Consumer Preference, Technological Forecasting and Social Change, 86, 2014.
[12] P. W. Murray, B. Agard and M. A. Barajas, Forecasting Supply Chain Demand by Clustering Customers, IFAC-PapersOnLine, 48(3), 2015.
[13] Y. Pang, B. Yao, X. Zhou, et al., Hierarchical Electricity Time Series Forecasting for Integrating Consumption Patterns Analysis and Aggregation Consistency, in Proceedings of the Twenty-Seventh International Joint Conference on Artificial Intelligence, 2018.
[14] M. A. Villegas, D. J. Pedregal and J. R. Trapero, A Support Vector Machine for Model Selection in Demand Forecasting Applications, Computers & Industrial Engineering, 121, 2018.
[15] H. Abbasimehr, M. Shabani and M. Yousefi, An Optimized Model Using LSTM Network for Demand Forecasting, Computers & Industrial Engineering, 143(3), 2020.
[16] M. Gaur, S. Goel and E. Jain, Comparison Between Nearest Neighbours and Bayesian Network for Demand Forecasting in Supply Chain Management, in 2nd International Conference on Computing for Sustainable Global Development (INDIACom), 2015.
[17] J. Wei, M. Bosma, V. Y. Zhao, et al., Finetuned Language Models Are Zero-Shot Learners, arXiv preprint arXiv:2109.01652., 2021.
[18] T. B. Brown, Language Models Are Few-Shot Learners., arXiv preprint arXiv:2005.14165, 2020.
[19] J. Wei, X. Wang, D. Schuurmans, et al., Chain-of-Thought Prompting Elicits Reasoning in Large Language Models, in Proceedings of the 36th International Conference on Neural Information Processing Systems (NIPS), 2024.
[20] Z. Zhang, A. Zhang, M. Li, et al., Automatic Chain of Thought Prompting in Large Language Models, arXiv preprint arXiv:2210.03493., 2022.
[21] P. Lewis, E. Perez, A. Piktus, et al., Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks, Advances in Neural Information Processing Systems, 33, 2020.
[22] N. Nassibi, H. Fasihuddin and L. Hsairi, A Proposed Demand Forecasting Model by Using Machine Learning for Food Industry, in ICFNDS '22: Proceedings of the 6th International Conference on Future Networks & Distributed Systems, 2023.
[23] S. S. Arnob, A. I. M. S. Arefin, A. Y. Saber, et al., Energy Demand Forecasting and Optimizing Electric Systems for Developing Countries, in IEEE Access, 11, 2023.
[24] V. Pilinkienė, Market Demand Forecasting Models and Their Elements in the Context of Competitive Market, Engineering Economics, 5(60), 2008.
[25] H. Jin, Y. Zhang, D. Meng, et al., A Comprehensive Survey on Process-Oriented Automatic Text Summarization with Exploration of LLM-Based Methods, arXiv preprint arXiv:2403.02901., 2024.
[26] H. Xue and F. D. Salim, Promptcast: A New Prompt-Based Learning Paradigm for Time Series Forecasting, IEEE Transactions on Knowledge and Data Engineering (Early Access), 2023. |
| Description: | 碩士
國立政治大學
資訊科學系碩士在職專班
110971019 |
| Source URI: | http://thesis.lib.nccu.edu.tw/record/#G0110971019 |
| Data Type: | thesis |
| Appears in Collections: | [資訊科學系碩士在職專班] 學位論文
|
Files in This Item:
| File |
Description |
Size | Format | |
|---|
| 101901.pdf | | 2431Kb | Adobe PDF | 0 | View/Open |
|
All items in 政大典藏 are protected by copyright, with all rights reserved.
|
