مقاله انگلیسی رایگان در مورد طراحی آزمایش جهت بهینه سازی یادگیری عمیق معماری برای پیش بینی سری زمانی غیر خطی – الزویر ۲۰۱۸

۱- Introduction

۲- Neural Network Model

۳- Experimental Plan

۴- Results and Discussion

۵- Conclusions and Future Works

۶- Acknowledgments

References

Abstract

The neural architecture is very substantial in order to construct a neural network model that produce a minimum error. Several factors among others include the input choice, the number of hidden layers, the series length, and the activation function. In this paper we present a design of experiment in order to optimize the neural network model. We conduct a simulation study by modeling the data generated from a nonlinear time series model, called subset 3 exponential smoothing transition auto-regressive (ESTAR ([3]). We explore a deep learning model, called deep feedforward network and we compare it to the single hidden layer feedforward neural network. Our experiment resulted in that the input choice is the most important factor in order to improve the forecast performance as well as the deep learning model is the promising approach for forecasting task.

Introduction

Time series is an observational data that is collected over time with the same time periods, such as in hours, days, weeks, months, and years [11]. Based on the data pattern, time series models are divided into two, namely linear time series model and nonlinear time series model. One very flexible method of forecasting time series data that contains both linear and nonlinear patterns is the neural network. The advantage of using a neural network is that it is not necessary to determine the shape of a particular model because the model is adaptively formed based on the features presented from the data [23]. Neural network adopts the biological neuron workings consisting of neurons as input processing, then the existing input value will be summed by a function of the summing function, and gives output based on the weight. Neural network models are widely applied in various fields of forecasting such as stock prices [10, 8, 14, 12, 6, 7], inflowoutflow [18], electricity consumption [4], and interest rates [19].