Abstract

Abstract

Abstract
Faced with the emergence of the Covid-19 pandemic, and to better understand and contain the disease's spread, health organisations increased the collaboration with other organisations sharing health data with data scientists and researchers. Data analysis assists such organisations in providing information that could help in decision-making processes. For this purpose, both national and regional health authorities provided health data for further processing and analysis. Shared data must comply with existing data protection and privacy regulations. Therefore, a robust de-identification procedure must be used, and a re-identification risk analysis should also be performed. De-identified data embodies state-of-the-art approaches in Data Protection by Design and Default because it requires the protection of direct and indirect identifiers (not just direct). This article highlights the importance of assessing re-identification risk before data disclosure by analysing a data set of individuals infected by Covid-19 that was made available for research purposes. We stress that it is highly important to make this data available for research purposes and that this process should be based on the state of the art methods in Data Protection by Design and by Default. Our main goal is to consider different re-identification risk analysis scenarios since the information on the intruder side is unknown. Our conclusions show that there is a risk of identity disclosure for all of the studied scenarios. For one, in particular, we proceed to an example of a re-identification attack. The outcome of such an attack reveals that it is possible to identify individuals with no much effort.

Abstract
The No Free Lunch (NFL) theorems have sparked intense debate since their publication, from theoretical and practical perspectives. However, to this date, no discussion is provided concerning its impact in the established field of imbalanced domain learning (IDL), known for its challenges regarding learning and evaluation processes. Most importantly, understanding the effect of commonly used solutions in such a field would prove very useful for future research. In this paper, we study the impact of data preprocessing methods, also known as resampling strategies, under the framework of the NFL theorems. Focusing on binary classification tasks, we claim that in IDL settings, when given a learning algorithm and a uniformly distributed set of target functions, the core conclusions of the NFL theorems are extensible to resampling strategies. As such, given no a priori knowledge or assumptions concerning data domains, any two resampling strategies are identical concerning their impact in the performance of predictive models. We provide a theoretical analysis and discussion on the intersection between IDL and the NFL theorems to support such a claim. Also, we collect empirical evidence via a thorough experimental study, including 98 data sets from multiple real-world knowledge domains.

Abstract
Extreme and rare events, such as spikes in air pollution or abnormal weather conditions, can have serious repercussions. Many of these sorts of events develop through spatio-temporal processes. Timely and accurate predictions are a most valuable tool in addressing their impact. We propose a new set of resampling strategies for imbalanced spatio-temporal forecasting tasks, which introduce bias into formerly random processes. This bias is a combination of a spatial and a temporal weight, which can be either static or relevance-aware, and includes a hyper-parameter that regulates the relative importance of the temporal and spatial dimensions in the selection of observations during under- or over-sampling. We test and compare our proposals against standard versions of the strategies on 10 different geo-referenced numeric time series, using 3 distinct off-the-shelf learning algorithms. Experimental results show that our proposals provide an advantage over random resampling strategies in imbalanced numerical spatio-temporal forecasting tasks.