Smart Data Representations: Impact on the Accuracy of Deep Neural Networks

TL;DR

This paper investigates how different data representations affect the accuracy of deep neural networks in energy time series forecasting, revealing that representation choice impacts performance variably across forecast horizons.

Contribution

It provides an analysis of the influence of data representations on neural network accuracy in energy forecasting, highlighting the importance of representation selection.

Findings

Data representations significantly influence forecasting accuracy.

Impact varies with forecast horizon.

Different representations can improve or impair performance.

Abstract

Deep Neural Networks are able to solve many complex tasks with less engineering effort and better performance. However, these networks often use data for training and evaluation without investigating its representation, i.e.~the form of the used data. In the present paper, we analyze the impact of data representations on the performance of Deep Neural Networks using energy time series forecasting. Based on an overview of exemplary data representations, we select four exemplary data representations and evaluate them using two different Deep Neural Network architectures and three forecasting horizons on real-world energy time series. The results show that, depending on the forecast horizon, the same data representations can have a positive or negative impact on the accuracy of Deep Neural Networks.