We study in this paper the consequences of using the Mean Absolute Percentage Error (MAPE) as a measure of quality for regression models. We show that finding the best model under the MAPE is equivalent to doing weighted Mean Absolute Error (MAE) regression. We show that universal consistency of Empirical Risk Minimization remains possible using the MAPE instead of the MAE.

This paper applies a value-based management framework to critically review empirical research in managerial accounting. This framework enables us to place the exceptionally diverse set of managerial accounting studies from the past several decades into an integrated structure. Our synthesis highlights the many consistent results in prior research, identifies remaining gaps and inconsistencies, ...

Methodological norms in economic theorising are interpreted as rational strategies to optimise some epistemic utility functions. A definition of ‘empirical verisimilitude’ is defended as a plausible interpretation of the epistemic preferences of researchers. Some salient differences between the scientific strategies of physics and of economics are derived from the comparison of the relative cos...

A generic fusion problem is studied for multiple sensors whose outputs are probabilistically related to their inputs according to unknown distributions. Sensor measurements are provided as iid input-output samples, and an empirical risk minimization method is described for designing fusers with distribution-free performance bounds. The special cases of isolation and projective fusers for classi...

The retrieval and preparation of public data on software development calls for more than just technical skills. In addition, care and judgement are needed to avoid disproportionate costs to the providers of data or unnecessary embarrassment to the participants tracked in the data. Taking the extraction of bug scenarios as a use case, we illustrate these concerns and discuss how they could be tr...

In this paper, we study a two-category classification problem. We indicate the categories by labels Y = 1 and Y = −1. We observe a covariate, or feature, X ∈ X ⊂ R. Consider a collection {ha} of classifiers indexed by a finite-dimensional parameter a, and the classifier ha∗ that minimizes the prediction error over this class. The parameter a∗ is estimated by the empirical risk minimizer ân over...

We address the problem of sequential prediction in the heteroscedastic setting, when both the signal and its variance are assumed to depend on explanatory variables. By applying regret minimization techniques, we devise an efficient online learning algorithm for the problem, without assuming that the error terms comply with a specific distribution. We show that our algorithm can be adjusted to ...

There is increasing support for sharing individual-level data generated by medical and public health research. This scoping review of empirical research and conceptual literature examined stakeholders' perspectives of ethical best practices in data sharing, particularly in low- and middle-income settings. Sixty-nine empirical and conceptual articles were reviewed, of which, only five were empir...

Climate change has a complex effect on snow at the regional scale. The change in snow patterns under climate change remains unknown for certain regions. Here, we used high spatiotemporal resolution snow-related variables simulated by a weather research and forecast model (WRF) including snowfall, snow water equivalent and snow depth along with fractional snow cover (FSC) data extracted from Mod...