The stock market plays an important role in the entire financial market and price of stock gets impacted by the performance of global indices and the world news. Several researchers used the sentiment from the news and technical indicators to predict the trends of a stock price. However, there is a limited number of studies used the performance of the global indicator and currency exchange rates. As we are living in a connected world, every country is dependent upon the other countries for the trades and services. Therefore, global indicator and currency exchange rates as a significant effect on the stock market indices. In this study, we propose an ensemble of Long Short-Term Memory (LSTM) model to predict the stock price using global indices, exchange rate, historical stock price, world news and technical indicators. Further, the analysis was extended to some of the benchmark classification models to compare the performance. The preliminary result shows an ensemble of three LSTM models yields an accuracy of 60% with optimum recall and true negative score, which shows the proposed model, predicts increasing and decreasing trends equally well and stable.
Volume 12 | 07-Special Issue
Pages: 1668-1676
DOI: 10.5373/JARDCS/V12SP7/20202273