V‐Awake: A Visual Analytics Approach for Correcting Sleep Predictions from Deep Learning Models

The usage of deep learning models for tagging input data has increased over the past years because of their accuracy and high‐performance. A successful application is to score sleep stages. In this scenario, models are trained to predict the sleep stages of individuals. Although their predictive accuracy is high, there are still mis classifications that prevent doctors from properly diagnosing sleep‐related disorders. This paper presents a system that allows users to explore the output of deep learning models in a real‐life scenario to spot and analyze faulty predictions. These can be corrected by users to generate a sequence of sleep stages to be examined by doctors. Our approach addresses a real‐life scenario with absence of ground truth. It differs from others in that our goal is not to improve the model itself, but to correct the predictions it provides. We demonstrate that our approach is effective in identifying faulty predictions and helping users to fix them in the proposed use case.     

Identifying potential areas of understorey coffee in Ethiopia’s highlands using predictive modelling

Coffee production is one of the main economic activities in Ethiopia, representing about 40% of the country’s economy. Coffee is particularly important in the Ethiopian highlands, where appropriate climate allows higher productivity and quality. The Ethiopian highlands also host an outstanding biodiversity, being considered one of the world’s most important biodiversity hotspots. In this context, conciliating agricultural practices with biodiversity conservation is a priority task for researchers and other stakeholders. However, identifying and mapping understorey coffee plantations in Ethiopian highlands is particularly challenging due to the presence of scattered exotic trees and the characteristics of understorey cultivation. In this research, we mapped potential areas of understorey coffee using predictive modelling and evaluated how projected changes in climate would affect the suitability of coffee production in the study area. Landscape maps, which were mapped using remote-sensing data based on object-based image analysis, remotely sensed spectral vegetation indices, and climatic variables were used to delineate probability maps showing the most likely location of understorey coffee. Normalized difference vegetation index and maximum temperature and precipitation were considered the best predictors for explaining the spatial distribution of understorey coffee. The accuracy of the probability map was validated based on existing understorey coffee areas mapped during field surveys. In addition, we show that potential changes in temperature and precipitation by 2050 are likely to shift suitable areas of understorey coffee to higher altitudes, affecting the landscape changes dynamics in the region.     

Response monitoring in children with phenylketonuria.

Phenylketonuria (PKU) is characterized by a disruption in the metabolism of phenylalanine and is associated with dopamine deficiency (Diamond, Prevor, Callender, & Druin, 1997) and cerebral white matter abnormalities (e.g., Anderson et al., 2007). From a neuropsychological perspective, prefrontal dysfunction is thought to underlie the deficits in executive abilities observed in individuals with PKU (Christ, Steiner, Grange, Abrams, & White, 2006; Diamond et al., 1997; White, Nortz, Mandernach, Huntington, & Steiner, 2001, 2002). The purpose of our study was to examine a specific aspect of executive ability, response monitoring, as measured by posterror slowing. The authors examined posterror reaction time (RT) in 24 children with well-controlled, early treated PKU and 25 typically developing control children using a go/no-go task. Results showed that RTs of both controls and children with PKU slowed significantly following the commission of errors. The magnitude of posterror slowing, however, was significantly less for children with PKU. These findings indicate deficient response monitoring in children with PKU. (PsycINFO Database Record (c) 2010 APA, all rights reserved)