本研究旨在探討並比較ARIMA(差分整合移動平均自我迴歸模型)模型與LSTM(長短期記憶)模型在國際恐慌事件期間對市場預測的績效。考慮到全球市場在面對重大政治、經濟或社會動蕩時的反應，本研究選取了具有代表性的國際恐慌事件作為研究對象，分析了這些事件對全球金融市場的影響。透過ARIMA模型，我們評估了線性時間序列分析在捕捉市場短期波動方面的能力。另一方面，LSTM模型則用於檢測市場對這些恐慌事件的長期和非線性反應。研究結果揭示了兩種模型在不同市場情況下的優缺點，並提供了在複雜市場環境下進行有效預測的見解。此外，本研究的發現對於金融分析師和投資者在理解市場動態以及制定應對策略時具有重要的實用意義。
This study aims to explore and compare the performance of ARIMA (Autoregressive Moving Average) and LSTM (Long Short-Term Memory) models in market predictions during international panic events. Recognizing the global market's response to major political, economic, or social upheavals, this research selected representative international panic events to analyze their impact on the global financial markets. Through the ARIMA model, we assessed the capability of linear time series analysis in capturing short-term market fluctuations. On the other hand, the LSTM model was utilized to detect long-term and non-linear responses of the market to these panic events. The results of the study reveal the strengths and weaknesses of both models under different market conditions and provide insights for effective forecasting in complex market environments. Additionally, the findings of this research hold practical significance for financial analysts and investors in understanding market dynamics and formulating response strategies.
