STR: Seasonal-Trend Decomposition Using Regression

We propose a new method for decomposing seasonal data: a seasonal-trend decomposition using regression (STR). Unlike other decomposition methods, STR allows for multiple seasonal and cyclic components, covariates, seasonal patterns that may have noninteger periods, and seasonality with complex topology. It can be used for time series with any regular time index, including hourly, daily, weekly, monthly, or quarterly data. It is competitive with existing methods when they exist and tackles many more decomposition problems than other methods allow. STR is based on a regularized optimization and so is somewhat related to ridge regression. Because it is based on a statistical model, we can easily compute confidence intervals for components, something that is not possible with most existing decomposition methods (such as seasonal-trend decomposition using Loess, X-12-ARIMA, SEATS-TRAMO, etc.). Our model is implemented in the R package stR, so it can be applied by anyone to their own data.

A novel hybrid deep learning model for taxi demand forecasting based on decomposition of time series and fusion of text data

Accurate taxi demand forecasting is significant to estimate the change of demand to further make informed decisions. Although deep learning methods have been widely applied for taxi demand forecasting, they neglect the complexity of taxi demand data and the impact of event occurrences, making it hard to effectively model the taxi demand in highly dynamic areas (e.g., areas with frequent event occurrences). Therefore, to achieve accurate and stable taxi demand forecasting in highly dynamic areas, a novel hybrid deep learning model is proposed in this study. First, to reduce the complexity of taxi demand time series, the seasonal-trend decomposition procedures based on loess is employed to decompose the time series into three simpler components (i.e., seasonal, trend, and remainder components). Then, different forecasting methods are adopted to handle different components to obtain robust forecasting results. Moreover, considering the instability and nonlinearity of the remainder component, this study proposed to fuse the event features (in particular, text data) to capture the unusual fluctuation patterns of remainder component and solve its extreme value problem. Finally, genetic algorithm is applied to determine the optimal weights for integrating the forecasting results of three components to obtain the final taxi demand. The experimental results demonstrate the better accuracy and reliability of the proposed model compared with other baseline forecasting models.

Contribution Analysis of “Suroboyo Bus” in Waste Management Based on Two Form of Complete Fourier Series Estimator

Plastic waste is a problem that almost exists in all countries. This problem arises because of the lack of facilities that can handle the plastic waste. Suroboyo Bus is an innovation for this problem because Suroboyo Bus uses plastic bottles as payment. The purpose of this research is to predict the percentage contribution of Suroboyo Bus in handling plastic waste. The Fourier series estimator performs well for data modeling with seasonal trend patterns. This paper examines two approaches to the Fourier series. The difference between the approaches is the inclusion of the phi (π) function in the model. The result shows the goodness of fit criterion model with π function are for and 0,08% for MAPE whereas the fit criterion model without π function is 100% for and 0,07% for MAPE. In conclusion, the Fourier series model without the π function is better because the Fourier series model without the π function is more satisfy the goodness of fit criteria than the Fourier series model with the π.