Abstract
Accurate evaluation of forecasting models is essential for ensuring reliable predictions. Current practices for evaluating and comparing forecasting models focus on summarising performance into a single score, using metrics such as SMAPE. We hypothesize that averaging performance over all samples dilutes relevant information about the relative performance of models. Particularly, conditions in which this relative performance is different than the overall accuracy. We address this limitation by proposing a novel framework for evaluating univariate time series forecasting models from multiple perspectives, such as one-step ahead forecasting versus multi-step ahead forecasting. We show the advantages of this framework by comparing a state-of-the-art deep learning approach with classical forecasting techniques. While classical methods (e.g. ARIMA) are long-standing approaches to forecasting, deep neural networks (e.g. NHITS) have recently shown state-of-the-art forecasting performance in benchmark datasets. We conducted extensive experiments that show NHITS generally performs best, but its superiority varies with forecasting conditions. For instance, concerning the forecasting horizon, NHITS only outperforms classical approaches for multi-step ahead forecasting. Another relevant insight is that, when dealing with anomalies, NHITS is outperformed by methods such as Theta. These findings highlight the importance of evaluating forecasts from multiple dimensions.

Abstract

Abstract

Abstract

Abstract
This research addresses the challenge of generating synthetic data that resembles real-world data while preserving privacy. With privacy laws protecting sensitive information such as healthcare data, accessing sufficient training data becomes difficult, resulting in an increased difficulty in training Machine Learning models and in overall worst models. Recently, there has been an increased interest in the usage of Generative Adversarial Networks (GAN) to generate synthetic data since they enable researchers to generate more data to train their models. GANs, however, may not be suitable for privacy-sensitive data since they have no concern for the privacy of the generated data. We propose modifying the known Conditional Tabular GAN (CTGAN) model by incorporating a privacy-aware loss function, thus resulting in the Private CTGAN (PCTGAN) method. Several experiments were carried out using 10 public domain classification datasets and comparing PCTGAN with CTGAN and the state-of-the-art privacy-preserving model, the Differential Privacy CTGAN (DP-CTGAN). The results demonstrated that PCTGAN enables users to fine-tune the privacy fidelity trade-off by leveraging parameters, as well as that if desired, a higher level of privacy.

Abstract
The importance of time series forecasting drives continuous research and the development of new approaches to tackle this problem. Typically, these methods are introduced through empirical studies that frequently claim superior accuracy for the proposed approaches. Nevertheless, concerns are rising about the reliability and generalizability of these results due to limitations in experimental setups. This paper addresses a critical limitation: the number and representativeness of the datasets used. We investigate the impact of dataset selection bias, particularly the practice of cherry-picking datasets, on the performance evaluation of forecasting methods. Through empirical analysis with a diverse set of benchmark datasets, our findings reveal that cherry-picking datasets can significantly distort the perceived performance of methods, often exaggerating their effectiveness. Furthermore, our results demonstrate that by selectively choosing just four datasets — what most studies report — 46% of methods could be deemed best in class, and 77% could rank within the top three. Additionally, recent deep learning-based approaches show high sensitivity to dataset selection, whereas classical methods exhibit greater robustness. Finally, our results indicate that, when empirically validating forecasting algorithms on a subset of the benchmarks, increasing the number of datasets tested from 3 to 6 reduces the risk of incorrectly identifying an algorithm as the best one by approximately 40%. Our study highlights the critical need for comprehensive evaluation frameworks that more accurately reflect real-world scenarios. Adopting such frameworks will ensure the development of robust and reliable forecasting methods. Copyright © 2025, Association for the Advancement of Artificia Intelligence (www.aaai.org). All rights reserved.