Search

Classic feature selection techniques remove those features that are either irrelevant or redundant, achieving a subset of relevant features that help to provide a better knowledge extraction. This allows the creation of compact models that are easier to interpret. Most of these techniques work over the whole dataset, but they are unable to provide the user with successful information when only instance information is needed. In short, given any example, classic feature selection algorithms do not give any information about which the most relevant information is, regarding this sample. This work aims to overcome this handicap by developing a novel feature selection method, called Saliency-based Feature Selection (SFS), based in deep-learning saliency techniques. Our experimental results will prove that this algorithm can be successfully used not only in Neural Networks, but also under any given architecture trained by using Gradient Descent techniques.

Socioemotional selectivity theory (SST) warrants the strong prediction that older adults respond more favorably to emotionally-meaningful versus knowledge-related appeals in persuasive messages, whereas younger adults lack this bias. However, potentially due to multivocality in conceptualizations and operationalizations of these appeals, previous studies found no uniform support for these age differences. Consequently, this article aims to provide a conceptualization and operationalization of emotionally-meaningful versus knowledge-related appeals that can be used in future research. The study consists of a conceptualization phase (literature review; expert meetings) and an operationalization phase (content analysis of persuasive messages). We developed a theoretically valid and reliable coding instrument, outlining three dimensions of emotionally-meaningful appeals (emotion regulation, optimizing the present, close social relationships) and three dimensions of knowledge-related appeals (knowledge acquisition, optimizing the future, novel social relationships). This instrument is intended to guide the selection and design of persuasive messages in effect studies that aim to test hypotheses derived from SST.

Ongoing advances in artificial intelligence (AI) are increasingly part of scientific efforts as well as the public debate and the media agenda, raising hopes and concerns about the impact of automated decision making across different sectors of our society. This topic is receiving increasing attention at both national and cross- national levels. The present report contributes to informing this public debate, providing the results of a survey with 958 participants recruited from high-quality sample of the Dutch population. It provides an overview of public knowledge, perceptions, hopes and concerns about the adoption of AI and ADM across different societal sectors in the Netherlands. This report is part of a research collaboration between the Universities of Amsterdam, Tilburg, Radboud, Utrecht and Eindhoven (TU/e) on automated decision making, and forms input to the groups’ research on fairness in automated decision making.

Previous studies have mainly focused on tailoring message content to match individual characteristics and preferences. Additional strategies, such as tailoring to individual preferences for the mode of information presentation, are proposed to increase message effectiveness. This study investigates the effect of a mode-tailored website compared to four non-tailored websites on younger and older adults’ attention and recall of information, employing a 5 (condition: tailored vs. text, text with illustrations, audiovisual, combination) (age: younger [25 – 45] vs. older [≥ 65] adults) design (N = 559). The mode-tailored condition (relative to non-tailored conditions) improved attention to the website and, consequently, recall in older adults, but not in younger adults. Younger adults recalled more from non-tailored information such as text only or text with illustrations, relative to tailored information.

This study proposes a theoretical framework for patient website functions and provides empirical input for the framework. A pilot survey among younger (aged 50 --- 64, M = 55.95, SD = 4.48, n = 21) and older (aged 65 --- 84, M = 72.79, SD = 6.33, n = 14) cancer patients revealed that patients' website preferences vary across type of website functions as well as across age groups. Whereas the majority of patients reported high preference for website functions in general (e.g., information provision), preferences varied across preferred delivery methods of these functions. Furthermore, differences in information preferences indicate a trend of younger patients preferring to search information themselves whereas older patients seem to prefer receiving disease relevant information and practical tips. We provide first evidence for patient preferences regarding specific website functions and thus provide practical implications for website design.

With the advent of extremely high dimensional datasets, dimensionality reduction techniques are becoming mandatory. Among many techniques, feature selection has been growing in interest as an important tool to identify relevant features on huge datasets --both in number of instances and features--. The purpose of this work is to demonstrate that standard feature selection methods can be parallelized in Big Data platforms like Apache Spark, boosting both performance and accuracy. We thus propose a distributed implementation of a generic feature selection framework which includes a wide group of well-known Information Theoretic methods. Experimental results on a wide set of real-world datasets show that our distributed framework is capable of dealing with ultra-high dimensional datasets as well as those with a huge number of samples in a short period of time, outperforming the sequential version in all the cases studied.

Eye tracking gives communication scholars the opportunity to move beyond self-reported measures by examining more precisely how much visual attention is paid to information. However, we lack insight into how eye-tracking data is used in communication research. This literature review provides an overview of how eye tracking is used in communication research by examining published articles from the top-25 ranked communication journals between 2005 and 2015. Our results showed that most eye-tracking research was employed in the field of advertising. Furthermore, most studies used eye tracking to measure (visual) attention and used this as the study’s dependent variable. A wide variety of eye-tracking measures were reported, including fixation time, fixation count, and visual shifts, and a wide variety of eye-tracking devices were used. Our results highlight opportunities for using eye tracking as well as identify other ways of using eye tracking to maximize its potential in communication research.

Background: Older adults are increasingly using the Internet for health information; however, they are often not able to correctly recall Web-based information (eHealth information). Recall of information is crucial for optimal health outcomes, such as adequate disease management and adherence to medical regimes. Combining effective message strategies may help to improve recall of eHealth information among older adults. Presenting information in an audiovisual format using conversational narration style is expected to optimize recall of information compared to other combinations of modality and narration style. Objective: The aim of this paper is to investigate the effect of modality and narration style on recall of health information, and whether there are differences between younger and older adults. Methods: We conducted a Web-based experiment using a 2 (modality: written vs audiovisual information) by 2 (narration style: formal vs conversational style) between-subjects design (N=440). Age was assessed in the questionnaire and included as a factor: younger (

In de praktijk van de Nederlandse gebiedsontwikkeling komt de nadruk steeds meer te liggen op de (her)ontwikkeling van binnenstedelijke gebieden. Terwijl publieke en private rollen verschuiven wordt de lange termijn belegger een centrale rol toegedicht. Uit dit onderzoek blijkt dat (Amerikaanse) belegger-gestuurde gebiedsontwikkeling perspectief biedt voor de in transitie verkerende Nederlandse praktijk. Het vraagt echter wel om een andere inrichting van zowel de investeringsstrategie van de belegger, als de ontwikkelstrategie voor het gebied. De belegger wordt geacht vanuit een multi-sectorale gebiedsgerichte aanpak een leidende rol op te nemen in het ontwikkelproces. Om met minder publiek kapitaal meer te kunnen bereiken zal de lokale overheid bereid moeten zijn ruimte te bieden aan privaat initiatief.

This study examines how adults pay attention to cognitive and affective illustrations on a cancer-related webpage and explores age-related differences in the attention to these cognitive and affective webpages. Results of an eye-tracking experiment (n = 20) showed that adults spent more time attending to the illustrations on the cognitive webpage than the illustrations on the affective webpage. Furthermore, older adults spent about 65% less time fixating the webpages than younger adults. Whereas older adults had less attention for illustrations on the cognitive webpage then younger adults, they spent equal time viewing the illustrations on the affective webpage as younger adults.

This book offers a coherent and comprehensive approach to feature subset selection in the scope of classification problems, explaining the foundations, real application problems and the challenges of feature selection for high-dimensional data. The authors first focus on the analysis and synthesis of feature selection algorithms, presenting a comprehensive review of basic concepts and experimental results of the most well-known algorithms. They then address different real scenarios with high-dimensional data, showing the use of feature selection algorithms in different contexts with different requirements and information: microarray data, intrusion detection, tear film lipid layer classification and cost-based features. The book then delves into the scenario of big dimension, paying attention to important problems under high-dimensional spaces, such as scalability, distributed processing and real-time processing, scenarios that open up new and interesting challenges for researchers. The book is useful for practitioners, researchers and graduate students in the areas of machine learning and data mining.

(Article in English) Le vécu inhérent à la prestation des soins aux personnes atteintes d’Alzheimer est relativement bien documenté.Toutefois,peu de recherches ont été réalisées du point de vue de la personne souffrant de démence.L’objectif de cette étude, qui s’inscrit dans une recherche qualitative sur les relations mères-filles en contexte de démence,notamment sur le processus de soins,est de recueillir les perceptions et les expériences des mères soignées par leur fille adulte.Guidés par des perspectives féministes et axées sur les étapes de vie,les chercheurs ont mené des entrevues en profondeur et semi-structurées auprès d’un échantillonnage diversifié de 10 femmes non-institutionnalisées et atteintes de troubles cognitifs légers à modérés.En général,les perceptions et les expériences des femmes sur le plan de la santé étaient façonnées par les rapports sociaux entre les sexes et par la manière dont ils sont construits.Bien que les mères rapportaient des interactions surtout positives avec leur fille,les idéologies culturelles de l’individualisme et des rapports axés sur la famille se traduisaient par des sentiments de «culpabilité-reconnaissance».Les participantes géraient leurs expériences contradictoires dans la situation où elles recevaient des soins de leur fille en donnant elles-mêmes des soins, en s’abstenant de demander des soins, en déterminant les soins et en acceptant les soins. Les auteurs recommandent l’apport de changements à l’échelle des pratiques,des politiques et de la recherche,qui doivent tenir compte de déterminants sociaux de la santé pertinents,comme les rapports sociaux entre les sexes et le soutien social. Ces changements contribueraient à la santé et au bien-être des femmes atteintes de démence. The caregiving experience within Alzheimer disease is fairly well documented. However,little research has been conducted from the perspective of the person living with dementia.The purpose of this study,part of a larger qualitative investigation of mother-daughter relationships within the care process of dementia, was to elicit the perceptions and experiences of mothers receiving care from their adult daughters.Guided by feminist and life-course perspectives,the researchers conducted in-depth,semi-structured interviews with a diverse sample of 10 community-dwelling women with mild to moderate cognitive impairment.In general,the health perceptions and experiences of the women were shaped by gender and how its meaning is constructed.While mothers reported mostly positive relationships with their daughters,cultural ideologies of individualism and familism manifested in feelings of “grateful guilt.” Participants managed their contradictory experiences of receiving care from their daughters by doing care,undemanding care,determining care, and accepting care. The authors recommend changes in practice,policy,and research,with the aim of addressing relevant social determinants of health such as gender and social support,thereby promoting the health and well-being of women with dementia.

Opinions regarding the appropriateness of elderly residents' sexual behaviors in a chronic care hospital and how to respond to inappropriate behaviors were surveyed. Study participation was open to all staff (N = 1,205), eligible residents (N = 182) and community-dwelling spouses (N = 103). Participation rates were 40% (residents), 42% (spouses), 34% (nursing staff), 50% (allied health staff), and 22% (support staff). Staff completed the questionnaire independently, while residents and spouses were offered self-completion or a structured interview. Almost all selected the interview. Residents and spouses were less tolerant than staff of residents' masturbating, engaging in sexual relationships, viewing sexual materials, and making sexual approaches to staff. Privacy was the primary determinant of appropriateness for behaviors for all groups. Staff and spouses were more likely to endorse counseling when behaviors were perceived as inappropriate than residents. Nurses endorsed counseling less frequently than allied health professionals and support staff. Nurses were more likely to have been approached sexually by a resident. Differences of opinion are interpreted in terms of cohort influences on values and contextual influences on behavior.