Speaker Invitat: Ionuț NODIȘ, University College London (UCL), London, United Kingdom
Titlul prezentării: De la Modele Tradiționale la Rețele Neuronale: Cum Schimbă AI-ul Predicția Piețelor Financiare?
Abstract: Forecasting stock returns with consistent accuracy is nearly improbable due to the stochastic nature of financial markets and their sensitivity to unpredictable macroeconomic events. However, forecasting volatility—a key component of risk management and options pricing—is more feasible, as volatility exhibits mean-reverting behavior and identifiable patterns. Traditional time series models such as ARIMA and GARCH have been widely used for financial forecasting, yet they often struggle to capture the complex, nonlinear relationships inherent in market data. Recent advancements in machine learning (ML) and deep learning (DL) have introduced alternative approaches that can enhance predictive performance. This study explores the effectiveness of hybrid models that integrate traditional forecasting techniques with deep learning methods, such as Long Short-Term Memory (LSTM) networks and Random Forest, to determine whether a combined approach improves predictive accuracy. While standalone deep learning models excel at capturing complex dependencies in data, they often lack interpretability and stability. By integrating them with traditional models, we assess whether hybrid approaches can provide a more reliable framework for forecasting volatility and, to a lesser extent, short-term price movements.
