An Experimental Study and Analysis of Long-Term Multi-Trending Trajectory Forecasting of Stock Indices Using Time Series Inferential Statistical Projection
| Autor: | G. Kavitha, S. Bhuvaneswari, D. Piriadarshani |
|---|---|
| Rok vydání: | 2022 |
| Předmět: | |
| Zdroj: | Remittances Review. 7:99-128 |
| ISSN: | 2059-6596 2059-6588 |
| DOI: | 10.47059/rr.v7i2.2415 |
| Popis: | The focus of this study is on time series inferential statistical projection and analysis for long-term multi-trending trajectory forecast modelling. Inferential statistics was used for projection and analysis of long-term multi-trending trajectory forecasting. The proposed model is trained, tested and validated on three different types of stocks viz. NVIDIA, UNG and IBM taken from the NASDAQ stock market index. In Phase I, the long-term deterministic multi-trending model is assessed and fitted. Model skewed residual series is assessed using the goodness of fit in Phase II. The future stock close prices and long-term multi-trending trajectory paths are simulated and analyzed in Phase III. Finally, in Phase IV, epsilon-skew-normal results and normal distribution assumptions are compared. The experimental result shows that an uptrend trajectory of historical close price simulates an uptrend trajectory in the forecasted close price, a downtrend trajectory of historical close price simulates a downtrend trajectory in the forecasted close price, a sideways or mixed trend trajectory of historical close price simulates a sideways or mixed trend trajectory in the forecasted close price. |
| Databáze: | OpenAIRE |
| Externí odkaz: |
|