Abstract
Saliency maps have been used as one possibility to interpret deep neural networks. This method estimates the relevance of each pixel in the image classification, with higher values representing pixels which contribute positively to classification.The goal of this study is to understand how the complexity of the network affects the interpretabilty of the saliency maps in classification tasks. To achieve that, we investigate how changes in the regularization affects the saliency maps produced, and their fidelity to the overall classification process of the network.The experimental setup consists in the calculation of the fidelity of five saliency map methods that were compare, applying them to models trained on the CIFAR-10 dataset, using different levels of weight decay on some or all the layers.Achieved results show that models with lower regularization are statistically (significance of 5%) more interpretable than the other models. Also, regularization applied only to the higher convolutional layers or fully-connected layers produce saliency maps with more fidelity. © 2020 IEEE.

Abstract
The prediction of overall survival in patients has an important role, especially in diseases with a high mortality rate. Encompassed in this reality, patients with oncological diseases, particularly the more frequent ones like woman breast cancer, can take advantage of a very good customization, which in some cases may even lead to a disease-free life. In order to achieve this customization, in this work a comparison between three algorithms (evolutionary, hierarchical and k-medoids) is proposed. After constructing a database with more than 800 breast cancer patients from a single oncology center with 15 clinical variables (heterogeneous data) and having 25% of the data missing, which illustrates a real clinical scenario, the algorithms were used to group similar patients into clusters. Using Tukey's HSD (Honestly Significant Difference) test, from both comparison between k-medoids and the other two approaches (evolutionary and hierarchical clustering) a statistical difference were detected (p- value < 0.0000001) as well as for the other comparison (evolutionary versus hierarchical clustering) - p-value = 0.0061354 - for a significance level of 95%. The future work will consist primarily in dealing with the missing data, in order to achieve better results in future prediction. © 2014, Springer International Publishing Switzerland.

Abstract
Brain-Computer Interface systems based on motor imagery are able to identify an individual's intent to initiate control through the classification of encephalography patterns. Correctly classifying such patterns is instrumental and strongly depends in a robust machine learning block that is able to properly process the features extracted from a subject's encephalograms. The main objective of this work is to provide an overall view on machine learning stages, aiming to answer the following question: "What are the steps in the classification process that we should worry about?". The obtained results suggest that future research in the field should focus on two main aspects: exploring techniques for dimensionality reduction, in particular, supervised linear approaches, and evaluating adequate validation schemes to allow a more precise interpretation of results.

Abstract

Abstract
The quality of mammography images is essential for the diagnosis of breast cancer and image imputation has become a popular technique to overcome noise, artifacts, and missing data to aid in the diagnosis of diseases. In this paper, we assess the performance of six imputation methodologies for the reconstruction of missing pixels in different morphologies in mammography images. The images included in this study are collected from four public datasets (CBIS-DDSM, Mini-MIAS, INbreast, and CSAW) and the imputation results are evaluated through the mean absolute error (MAE) and structural similarity index measure (SSIM). This study goes beyond the traditional evaluation of imputation algorithms, analyzing imputation quality, morphology preservation and classification performance. The effects of imputation on the morphology of cancer lesions are of utmost importance since it lays the foundation for physicians to interpret and analyze the imputation results. The results show that DIP is the most promising methodology for higher missing pixel rates, morphology preservation, and classifying malignant and benign images.

Abstract
With the increased use of machine learning algorithms to make decisions which impact people's lives, it is of extreme importance to ensure that predictions do not prejudice subgroups of the population with respect to sensitive attributes such as race or gender. Discrimination occurs when the probability of a positive outcome changes across privileged and unprivileged groups defined by the sensitive attributes. It has been shown that this bias can be originated from imbalanced data contexts where one of the classes contains a much smaller number of instances than the other classes. It is also important to identify the nature of the imbalanced data, including the characteristics of the minority classes' distribution. This paper presents FAWOS: a Fairness-Aware oversampling algorithm which aims to attenuate unfair treatment by handling sensitive attributes' imbalance. We categorize different types of datapoints according to their local neighbourhood with respect to the sensitive attributes, identifying which are more difficult to learn by the classifiers. In order to balance the dataset, FAWOS oversamples the training data by creating new synthetic datapoints using the different types of datapoints identified. We test the impact of FAWOS on different learning classifiers and analyze which can better handle sensitive attribute imbalance. Empirically, we observe that this algorithm can effectively increase the fairness results of the classifiers while not neglecting the classification performance. Source code can be found at: https://github.com/teresalazar13/FAWOS