Search

Inspired by the successful use of deep learning in computer vision, in this paper we introduce ForCNN , a novel deep learning method for univariate time series forecasting that mixes convolutional and dense layers in a single neural network. Instead of using conventional, numeric representations of time series data as input to the network, the proposed method considers visual representations of it in the form of images to directly produce point forecasts. Three variants of deep convolutional neural networks are examined to process the images, the first based on VGG-19, the second on ResNet-50, while the third on a self-designed architecture. The performance of the proposed approach is evaluated using time series of the M3 and M4 forecasting competitions. Our results suggest that image-based time series forecasting methods can outperform both standard and state-of-the-art forecasting models. • An image-based deep learning method for time series forecasting is proposed. • Numeric time series data are converted to images using simple line plots. • Deep convolutional neural networks are employed for producing forecasts from each image. • The proposed, image-based, method outperforms state-of-the-art forecasting methods. [ABSTRACT FROM AUTHOR]

• Empirical evaluation of data augmentation techniques for univariate time series forecasting. • Data augmentation techniques based on combining or upsampling the original time series improve the forecasting performance of neural networks. • Accuracy improvements from data augmentation are more significant when deeper neural networks are used. • Accuracy improvements from data augmentation become less significant as the initial size of the set increases. Neural networks have been proven particularly accurate in univariate time series forecasting settings, requiring however a significant number of training samples to be effectively trained. In machine learning applications where available data are limited, data augmentation techniques have been successfully used to generate synthetic data that resemble and complement the original train set. Since the potential of data augmentation has been largely neglected in univariate time series forecasting, in this study we investigate nine data augmentation techniques, ranging from simple transformations and adjustments to sophisticated generative models and a novel upsampling approach. We empirically evaluate the impact of data augmentation on forecasting accuracy considering both shallow and deep feed-forward neural networks and time series data sets of different sizes from the M4 and the Tourism competitions. Our results suggest that certain data augmentation techniques that build on upsampling and time series combinations can improve forecasting performance, especially when deep networks are used. However, these improvements become less significant as the initial size of the train set increases. [ABSTRACT FROM AUTHOR]

• Energy forecasting at a building level constitutes a cross-sectional hierarchy. • We employ Multiple Temporal Aggregation (MTA) to mitigate modeling uncertainty. • We propose a modification of MTA for high frequency time series. • We introduce a framework to combine MTA with cross-sectional hierarchies. • Empirical evaluation demonstrates substantial gains over conventional forecasting. Achieving high accuracy in energy consumption forecasting is critical for improving energy management and planning. However, this requires the selection of appropriate forecasting models, able to capture the individual characteristics of the series to be predicted, which is a task that involves a lot of uncertainty. When hierarchies of load from different sources are considered together, the uncertainty and complexity increase further. For example, when forecasting both at system and region level, not only the model selection problem is expanded to multiple time series, but we also require aggregation consistency of the forecasts across levels. Although hierarchical forecasting, such as the bottom-up, the top-down, and the optimal reconciliation methods, can address the aggregation consistency concerns, it does not resolve the model selection uncertainty. To address this issue, we rely on Multiple Temporal Aggregation (MTA), which has been shown to mitigate the model selection problem for low-frequency time series. We propose a modification of the Multiple Aggregation Prediction Algorithm, a special implementation of MTA, for high-frequency time series to better handle the undesirable effect of seasonality shrinkage that MTA implies and combine it with conventional cross-sectional hierarchical forecasting. The impact of incorporating temporal aggregation in hierarchical forecasting is empirically assessed using a real data set from five bank branches. We show that the proposed MTA approach, combined with the optimal reconciliation method, demonstrates superior accuracy, aggregation consistency, and reliable automatic forecasting. [ABSTRACT FROM AUTHOR]