Financial Time Series Data Forecasting for Event Handling

The volatile nature of the financial market makes it difficult to develop a reliable forecasting system for financial time series data, but it becomes unreliable when the financial market fluctuates unexpectedly due to events such as demonetization, earthquakes, or other government financial decisions. Financial market analysis based on various events is very crucial task due to drastic ups and down and more error prone especially for next day ahead forecasting. Financial time series data are affected by various factors and events, some of which influence prices of financial market directly and others that do so indirectly. Financial time series data can be affected by two different types of events: expected and unexpected events. Expected events are those which are almost known to everyone, such as a presidential election. On the other hand, an unexpected event occurs suddenly and without warning, such as natural disasters, wars, demonetization and other political activities. This study focuses on the creation of a resilient financial time series forecasting system based on ANN approaches.