Scenario Based Time Series Prediction Using Seasonal ARIMA

Rapidly evolving businesses generate massive amounts of time-stamped data sequences and cause a demand for both univariate and multivariate time series forecasting. For such data, traditional predictive models based on autoregression are often not sufficient to capture complex nonlinear relationships between multidimensional features and the time series outputs. An architecture for combined, multilevel time series prediction is proposed, which is suitable for many different universal regressors and combination methods. This architecture did not perform the results which are scenario based. Due to this the predicted result was not correct. Therefore an algorithm called seasonal ARIMA is used to predict the weather conditions where the seasonal and periodic item sets are taken. Regression and combination will be performed according to the threshold value, finally clusters are formed which will show the predicted result. The proposed system is evaluated in different scenarios and showed a clear prediction performance gain