定價策略在商業業務管理中扮演極重要的角色，越來越多的企業渴望更快速地做出最符合市場的決策，而隨著人工智慧與機器學習風潮興起，業界開始關注如何運用人工智慧與機器學習建立準確且自動化的價格預測系統。
價格的波動性，在市場交易面上格外被大家重視，價格變動性相對大的產業在價格的制訂上勢必得格外謹慎，而半導體產業則屬於價格波動性相對大的產業。在半導體產業中，各家公司的定價策略就顯得十分之重要，本研究以記憶體價格為例。
本研究之目的是透過機器學習演算法，開發更精準的自動化價格預測模型，而本研究提出之模型主要是運用一種機器學習模型―多層感知機（MLP Model）來進行模型的訓練，並加入十個產品共160天的歷史價格、四個具指標性之股市資訊、以及半導體產業相關新聞三個面向市場資訊，藉此建置四個價格修正模型來改善預測結果。機器學習訓練出合適的模型特徵和調整最佳參數，透過本研究提出之修正模型，達到修正時間序列SMA模型的效果，提供更精準的價格預測，以執行更符合市場的訂價策略。
從研究結果發現，對於DRAM產品線，模型一的模型修正成功率平均為57.04%；模型二的模型修正成功率平均為50.37%；模型三的模型修正成功率平均為50.37%；模型四的模型修正成功率平均為55.56%。而NAND Flash產品線，模型一的模型修正成功率平均為8.15%；模型二的模型修正成功率平均為6.67%；模型三的模型修正成功率平均為7.41%；模型四的模型修正成功率平均為8.15%。整體而言，模型修正成功率越高，MAPE下降率也會越大。
針對價格波動性較大的階段，研究結果不僅表明機器學習模型可做到記憶體的價格預測，且透過加入多種類型的市場資訊，將更能夠改善價格預測的精準度，可以提供定價策略的決策者一個準確且客觀的參考。

With the rise of artificial intelligence (AI) and machine learning (ML), more and more industries have been putting focuses on how to implement AI and ML to build accurate and automated price prediction systems.
Price volatility is particularly important on the trading side of the market. The industries with relatively high price volatility tend to be extra cautious in price setting, and the semiconductor industry is one of these industries. In the semiconductor field, the pricing strategy of each company is very important. As a result, this study takes memory pricing as example.
The objective of this research is to develop more accurate automated price prediction models based on machine learning algorithms, and the propose model is the Multi-Layer Perceptron (MLP) model. The model is trained by 160 days of historical prices for ten products, four indicative stock information, and three market-oriented information relating to semiconductor industry news. The three market-oriented information are added to build four price correction models in order to improve the prediction results. The method is able to train the appropriate features and adjust the optimal parameters to achieve the effect of correcting the time-series SMA model. By this result, one can provide a more accurate price prediction and execute a more market-compatible pricing strategy.
From the research results, for DRAM product line, the average success rate of model correction is 57.04% by Model 1, 50.37% by Model 2, 50.37% by Model 3, and 55.56% by Model 4. As for NAND Flash product line, the average success rate of model correction is 8.15% by Model 1, 6.67% by Model 2, 7.41% by Model 3, and 8.15% by Model 4. Overall, the higher the success rate of model correction, the greater the decline rate of MAPE.
The results show that the proper machine learning model can achieve memory price prediction during the period of high price volatility. Moreover, by adding multiple types of market information, the model can provide a more accurate and objective reference for decision makers on pricing strategies.

Ahn, J., & Lee, O. (2011). A Neural Network-based Semiconductor Price Prediction System.
Ausloos, M., Zhang, Y., & Dhesi, G. (2020). Stock index futures trading impact on spot price volatility. The CSI 300 studied with a TGARCH model. Expert Systems with Applications, 160, 113688.
Berger, J., Humphreys, A., Ludwig, S., Moe, W. W., Netzer, O., & Schweidel, D. A. (2020). Uniting the tribes: Using text for marketing insight. Journal of Marketing, 84(1), 1-25.
Bollen, J., Mao, H., & Pepe, A. (2010, October). Determining the Public Mood State by Analysis of Microblogging Posts. In ALIFE (pp. 667-668).
Bousqaoui, H., Achchab, S., & Tikito, K. (2017, October). Machine learning applications in supply chains: An emphasis on neural network applications. In 2017 3rd International Conference of Cloud Computing Technologies and Applications (CloudTech) (pp. 1-7). IEEE.
Brea, C., Anderson, J., Bartling, R., Mayo, R. & Mueller, F. (2018). Demand Forecasting with Advanced Analytics, retrieved from https://www.bain.com/insights/demandforecasting-with-advanced-analytics/
Chen, W., Zhang, Y., Yeo, C. K., Lau, C. T., & Lee, B. S. (2017, September). Stock market prediction using neural network through news on online social networks. In 2017 international smart cities conference (ISC2) (pp. 1-6). IEEE.
Eddy, A. R., & Saunders, G. B. (1980). New product announcements and stock prices. Decision sciences, 11(1), 90-97.
Flamm, K. (2003). Moore's law and the economics of semiconductor price trends. International Journal of Technology, Policy and Management, 3(2), 127-141.
Furbush, D. (1989). Program trading and price movement: Evidence from the October 1987 market crash. Financial Management, 68-83.
Gandomi, A., & Haider, M. (2015). Beyond the hype: Big data concepts, methods, and analytics. International journal of information management, 35(2), 137-144.
Harris, L. (1989). S&P 500 Cash Stock Price Volatilities. The Journal of Finance, 44(5), 1155–1175.
Hernández-Álvarez, M., Hernández, E. A. T., & Yoo, S. G. (2019, February). Stock Market Data Prediction Using Machine Learning Techniques. In International Conference on Information Technology & Systems (pp. 539-547). Springer, Cham.
Hoseinzade, E., & Haratizadeh, S. (2019). CNNpred: CNN-based stock market prediction using a diverse set of variables. Expert Systems with Applications, 129, 273-285.
Jiang, P. (2020) Review 2019 and Outlook 2020: Global and Taiwan Semiconductor Industry. IEK Topics. Retrieved from https://ieknet.iek.org.tw/iekrpt/rpt_more.aspx?actiontype=rpt&indu_idno=1&domain=20&rpt_idno=14452475 (Mar. 12. 2020)
Kadlec, P., & Gabrys, B. (2010, July). Adaptive on-line prediction soft sensing without historical data. In The 2010 international joint conference on neural networks (IJCNN) (pp. 1-8). IEEE.
Karabulut, Y. (2013, October). Can facebook predict stock market activity? In AFA 2013 San Diego Meetings Paper.
Lan, P. C., Kung, W. L., Ou, Y. L., Lin, C. Y., Hu, W. C., & Wang, Y. H. (2019, December). Machine learning model with technical analysis for stock price prediction: Empirical study of Semiconductor Company in Taiwan. In 2019 International Symposium on Intelligent Signal Processing and Communication Systems (ISPACS) (pp. 1-2). IEEE.
Liu, B. (2011). Web data mining: exploring hyperlinks, contents, and usage data (Vol. 1). Berlin: springer.
Mohan, S., Mullapudi, S., Sammeta, S., Vijayvergia, P., & Anastasiu, D. C. (2019, April). Stock price prediction using news sentiment analysis. In 2019 IEEE Fifth International Conference on Big Data Computing Service and Applications (BigDataService) (pp. 205-208). IEEE.
Nassirtoussi, A. K., Aghabozorgi, S., Wah, T. Y., & Ngo, D. C. L. (2014). Text mining for market prediction: A systematic review. Expert Systems with Applications, 41(16), 7653-7670.
Rasekhschaffe, K. C., & Jones, R. C. (2019). Machine learning for stock selection. Financial Analysts Journal, 75(3), 70-88.
Selvin, S., Vinayakumar, R., Gopalakrishnan, E. A., Menon, V. K., & Soman, K. P. (2017, September). Stock price prediction using LSTM, RNN and CNN-sliding window model. In 2017 international conference on advances in computing, communications and informatics (icacci) (pp. 1643-1647). IEEE.
Siami-Namini, S., Tavakoli, N., & Namin, A. S. (2018, December). A comparison of ARIMA and LSTM in forecasting time series. In 2018 17th IEEE International Conference on Machine Learning and Applications (ICMLA) (pp. 1394-1401). IEEE.
Stock-ai (2021). Global Semiconductor Market. Retrieved from https://stock-ai.com/semiconductor-billings
Stoll, H. R., & Whaley, R. E. (1988). Volatility and futures: Message versus messenger. Journal of Portfolio Management, 14(2), 20.
Tetlock, P. C. (2007). Giving content to investor sentiment: The role of media in the stock market. The Journal of finance, 62(3), 1139-1168.
Wang, S., Luo, M. A. Y., Jussa, J., Wang, A., Rohal, G., & December, D. E. (2014). Signal processing: The rise of the machines, iii. Deutsche Bank Quantitative Strategy.
Wen, M., Li, P., Zhang, L., & Chen, Y. (2019). Stock market trend prediction using high-order information of time series. Ieee Access, 7, 28299-28308.