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.

Machine learning has been increasingly applied to solve forecasting problems. Classical forecasting approaches, such as ARIMA or exponential smoothing are being replaced by machine learning regression algorithms, such as XGBoost, Gaussian processes or deep learning. However, despite the increasing attention, there are still doubts about the forecasting performance of machine learning methods. Makridakis, one of the most prominent names in the forecasting literature, has recently presented evidence that classical methods systematically outperform machine learning approaches for univariate time series forecasting [1]. This includes algorithms such as the LSTM, multi-layer perceptron or Gaussian processes.

More specifically, Professor Makridakis' proposal on "Clustered Ensemble of Specialist Amazon GluonTS Models for Time Series Forecasting" was chosen to be funded by the AWS Machine Learning Research Awards program, through which Professor Makridakis will be conducting research into improving Amazon's forecasting software and, ultimately, forecasting accuracy, using Machine/Deep Learning (a form of AI). Amazon has been an invaluable partner to Professor Makridakis and to UNIC, sponsoring the M4 Forecasting Competition in 2018. The M (or Makridakis) Competitions are the brainchild of Professor Makridakis, an initiative spanning almost four decades, with iterations taking place in 1982 (M1), 1993 (M2), 2000 (M3) and 2018 (M4). The current M5 Competition, launched in partnership with Walmart and Kaggle, is well underway (running till 30 June 2020) with 25,000 thousand participants to date. Commenting on the award, Professor Makridakis, the Director of the Makridakis Open Forecasting Center (MOFC) at UNIC, thanked Amazon for their trust and support.

"Over the last decade a new technology has begun to take hold in... business, one so new that its significance is still difficult to evaluate. While many aspects of this technology are uncertain, it seems clear that it will move into the managerial scene rapidly, with definite and far reaching impact on managerial organization." This article examines the near-term impact of expert system technology on work and the organization. First, an approach is taken for forecasting the likely extent of the diffusion, or success, of the technology. Next, the case of advanced manufacturing technologies and their effects is considered.

This article examines the near-term impact of expert system technology on work and the organization. First, an approach is taken for forecasting the likely extent of the diffusion, or success, of the technology. Next, the case of advanced manufacturing technologies and their effects is considered. From this analysis, a framework is constructed for viewing the impact of these technologies -- and technologies in general -- as a function of the technology itself; market realities; and personal, organizational, and societal values and policy choices. Two scenarios are proposed with respect to the application of this framework to expert systems. The first concludes that expert systems will have little impact on the nature of work and the organization. The second scenario posits that expert system diffusion will be pulled by, and will be a contributing factor toward, the evolution of the lean, flexible, knowledge-intensive, postindustrial organization.