Abstract

Abstract
[No abstract available]

Abstract
In this paper we introduce and develop the concept of interval-weighted networks (IWN), a novel approach in Social Network Analysis, where the edge weights are represented by closed intervals composed with precise information, comprehending intrinsic variability. We extend IWN for both Newman's modularity and modularity gain and the Louvain algorithm, considering a tabular representation of networks by contingency tables. We apply our methodology to two real-world IWN. The first is a commuter network in mainland Portugal, between the twenty three NUTS 3 Regions (IWCN). The second focuses on annual merchandise trade between 28 European countries, from 2003 to 2015 (IWTN). The optimal partition of geographic locations (regions or countries) is developed and compared using two new different approaches, designated as Classic Louvain and Hybrid Louvain , which allow taking into account the variability observed in the original network, thereby minimizing the loss of information present in the raw data. Our findings suggest the division of the twenty three Portuguese regions in three main communities for the IWCN and between two to three country communities for the IWTN. However, we find different geographical partitions according to the community detection methodology used. This analysis can be useful in many real-world applications, since it takes into account that the weights may vary within the ranges, rather than being constant.

Abstract
Concerns with the interpretability of ML models are growing as the technology is used in increasingly sensitive domains (e.g., health and public administration). Synthetic data can be used to understand models better, for instance, if the examples are generated close to the frontier between classes. However, data augmentation techniques, such as Generative Adversarial Networks (GAN), have been mostly used to generate training data that leads to better models. We propose a variation of GANs that, given a model, generates realistic data that is classified with low confidence by a given classifier. The generated examples can be used in order to gain insights on the frontier between classes. We empirically evaluate our approach on two well-known image classification benchmark datasets, MNIST and Fashion MNIST. Results show that the approach is able to generate images that are closer to the frontier when compared to the original ones, but still realistic. Manual inspection confirms that some of those images are confusing even for humans.

Abstract
Evaluating predictive models is a crucial task in predictive analytics. This process is especially challenging with time series data because observations are not independent. Several studies have analyzed how different performance estimation methods compare with each other for approximating the true loss incurred by a given forecasting model. However, these studies do not address how the estimators behave for model selection: the ability to select the best solution among a set of alternatives. This paper addresses this issue. The goal of this work is to compare a set of estimation methods for model selection in time series forecasting tasks. This objective is split into two main questions: (i) analyze how often a given estimation method selects the best possible model; and (ii) analyze what is the performance loss when the best model is not selected. Experiments were carried out using a case study that contains 3111 time series. The accuracy of the estimators for selecting the best solution is low, despite being significantly better than random selection. Moreover, the overall forecasting performance loss associated with the model selection process ranges from 0.28 to 0.58%. Yet, no considerable differences between different approaches were found. Besides, the sample size of the time series is an important factor in the relative performance of the estimators.

Abstract
The early detection of anomalous events in time series data is essential in many domains of application. In this paper we deal with critical health events, which represent a significant cause of mortality in intensive care units of hospitals. The timely prediction of these events is crucial for mitigating their consequences and improving healthcare. One of the most common approaches to tackle early anomaly detection problems is through standard classification methods. In this paper we propose a novel method that uses a layered learning architecture to address these tasks. One key contribution of our work is the idea of pre-conditional events, which denote arbitrary but computable relaxed versions of the event of interest. We leverage this idea to break the original problem into two hierarchical layers, which we hypothesize are easier to solve. The results suggest that the proposed approach leads to a better performance relative to state of the art approaches for critical health episode prediction.