A0864
Title: Bridging search behavior and market dynamics: A hybrid model for high-frequency financial data
Authors: Soudeep Deb - Indian Institute of Management Bangalore (India) [presenting]
Archi Roy - Indian Institute of Science Education and Research Pune (India)
Anitha Pathlavath - Indian Institute of Management Bangalore (India)
Abstract: With the rapid evolution of financial markets, cryptocurrency has emerged as a unique and highly volatile asset class. The purpose is to examine how Google search volumes related to Bitcoin can serve as indicators of market sentiment and aid in forecasting price movements. Given the non-linearity and temporal dependencies in high-frequency financial data, a hybrid model combining non-parametric regression (NR) with long short-term memory (LSTM) networks is proposed. The NR-LSTM model outperforms traditional approaches, effectively capturing complex patterns and highlighting the value of search data in predicting Bitcoin price dynamics. These findings contribute to advancing cryptocurrency forecasting methodologies.
