Neural networks and traditional time series methods: a synergistic combination in state economic forecasts

Abstract

Ever since the initial planning for the 1997 Utah legislative session, neural-network forecasting techniques have provided valuable insights for analysts forecasting tax revenues. These revenue estimates are critically important since agency budgets, support for education, and improvements to infrastructure all depend on their accuracy. Underforecasting generates windfalls that concern taxpayers, whereas overforecasting produces budget shortfalls that cause inadequately funded commitments. The pattern finding ability of neural networks gives insightful and alternative views of the seasonal and cyclical components commonly found in economic time series data. Two applications of neural networks to revenue forecasting clearly demonstrate how these models complement traditional time series techniques. In the first, preoccupation with a potential downturn in the economy distracts analysis based on traditional time series methods so that it overlooks an emerging new phenomenon in the data. In this case, neural networks identify the new pattern that then allows modification of the time series models and finally gives more accurate forecasts. In the second application, data structure found by traditional statistical tools allows analysts to provide neural networks with important information that the networks then use to create more accurate models. In summary, for the Utah revenue outlook, the insights that result from a portfolio of forecasts that includes neural networks exceeds the understanding generated from strictly statistical forecasting techniques. In this case, the synergy clearly results in the whole of the portfolio of forecasts being more accurate than the sum of the individual parts.

Improvement of borrowing channel assignment for patterned traffic load by online cellular probabilistic self-organizing map

“…Owing to different nature of problems, different types of networks were designed and used accordingly. These include multi-layer back-propagation (BP) neural network [31][32], radial basis function (RBF) neural network [33][34], time delay neural network [35][36], recurrent neural network (RNN) [37][38][39]44], support vector machine (SVM) [40][41], nonlinear autoregressive integrated neural network (NARI) [42][43], etc. Among them, BP, RBF, NARI neural networks and SVM have feed-forward architectures such that they are suitable for stationary data sets.…”

“…Neural networks have been widely used in the prediction problem [29][30][31][32][33][34][35][36][37][38][39][40][41][42][43][44]. Without a priori assumption about the properties of the data, neural network can approximate any nonlinear function universally [30].…”

Aproximaciones neuronales univariantes para la predicción de caudales diarios en cuencas portuguesas

“…De este modo, se evaluó el funcionamiento de modelos de redes neuronales de retropropagación entrenadas con el algoritmo Levenberg-Marquard (Shepherd, 1997) para la predicción a corto plazo de caudales diarios. Asimismo para comprobar si mejores resultados podían conseguirse, dos variaciones de los modelos típicos de RNCs se probaron: a) Se realizó un proceso de convolución de las variables de entrada al modelo de RNC (De Vries y Principe, 1991); y b) Se desarrolló una metodología híbrida combinando RNC y modelos ARIMA para aprovechar las cualidades de ambas metodologías en el modelado no lineal y lineal, respectivamente (Wedding y Cios, 1996;Hansen y Nelson, 1997;Zhang, 2003).…”

An Adaptive Approach to Market Development Forecasting

“…The adaptive nature of neural networks enhances their capability in time series prediction. However, effective neural network models for time series forecasting often require cross-section factors [10], or incorporate specific heuristics [11], or combine with traditional time series methods [12].…”