Abstract:
This thesis explores the use of neural networks and Long Short-Term Memory
(LSTM) techniques for stock prediction. The focus is on understanding the potential
of these techniques in forecasting stock prices by analyzing financial data and
developing predictive models. The study will evaluate the effectiveness of LSTM
models in comparison to traditional time-series models and assess the impact of
different hyper parameters on the accuracy of stock predictions. The aim is to
provide insights into the advantages and limitations of using LSTM and Neural
networks for stock prediction and to provide guidelines for practitioners and
researchers in the field.
