Active Learning of Very-High Resolution Optical Imagery with SVM: Entropy vs Margin Sampling

An active learning method is proposed for the semi-automatic selection of training sets in remote sensing image classification. The method adds iteratively to the current training set the unlabeled pixels for which the prediction of an ensemble of classifiers based on bagged training sets show maximum entropy. This way, the algorithm selects the pixels that are the most uncertain and that will improve the model if added in the training set. The user is asked to label such pixels at each iteration. Experiments using support vector machines (SVM) on an 8 classes QuickBird image show the excellent performances of the methods, that equals accuracies of both a model trained with ten times more pixels and a model whose training set has been built using a state-of-the-art SVM specific active learning method

Regulation of foreign gene in fowlpox virus by a vaccinia virus promoter.

A vaccinia virus promoter was evaluated for regulation of a foreign gene in fowlpox virus by a transient expression assay. Fowlpox virus-infected quail cells, transfected with plasmid DNA containing chloramphenicol acetyltransferase (CAT) gene ligated to a vaccinia virus promoter, expressed CAT activity. No CAT activity was detected either in uninfected cells or fowlpox virus-infected cells. These results indicated that a heterologous vaccinia virus promoter can regulate expression of a foreign gene in fowlpox virus.

Olfactory cues potentiate learning of distant visuospatial information

The influence of proximal olfactory cues on place learning and memory was tested in two different spatial tasks. Rats were trained to find a hole leading to their home cage or a single food source in an array of petri dishes. The two apparatuses differed both by the type of reinforcement (return to the home cage or food reward) and the local characteristics of the goal (masked holes or salient dishes). In both cases, the goal was in a fixed location relative to distant visual landmarks and could be marked by a local olfactory cue. Thus, the position of the goal was defined by two sets of redundant cues, each of which was sufficient to allow the discrimination of the goal location. These experiments were conducted with two strains of hooded rats (Long-Evans and PVG), which show different speeds of acquisition in place learning tasks. They revealed that the presence of an olfactory cue marking the goal facilitated learning of its location and that the facilitation persisted after the removal of the cue. Thus, the proximal olfactory cue appeared to potentiate learning and memory of the goal location relative to distant environmental cues. This facilitating effect was only detected when the expression of spatial memory was not already optimal, i.e., during the early phase of acquisition. It was not limited to a particular strain.

Multitask learning of environmental spatial data

The present research deals with an application of artificial neural networks for multitask learning from spatial environmental data. The real case study (sediments contamination of Geneva Lake) consists of 8 pollutants. There are different relationships between these variables, from linear correlations to strong nonlinear dependencies. The main idea is to construct a subsets of pollutants which can be efficiently modeled together within the multitask framework. The proposed two-step approach is based on: 1) the criterion of nonlinear predictability of each variable ?k? by analyzing all possible models composed from the rest of the variables by using a General Regression Neural Network (GRNN) as a model; 2) a multitask learning of the best model using multilayer perceptron and spatial predictions. The results of the study are analyzed using both machine learning and geostatistical tools.

Well, you have hepatic metastases: Use of technical language by medical students in simulated patient interviews.

OBJECTIVE: This research explored medical students' use and perception of technical language in a practical training setting to enhance skills in breaking bad news in oncology. METHODS: Terms potentially confusing to laypeople were selected from 108 videotaped interviews conducted in an undergraduate Communication Skills Training. A subset of these terms was included in a questionnaire completed by students (N=111) with the aim of gaining insight into their perceptions of different speech registers and of patient understanding. Excerpts of interviews were analyzed qualitatively to investigate students' communication strategies with respect to these technical terms. RESULTS: Fewer than half of the terms were clarified. Students checked for simulated patients' understanding of the terms palliative and metastasis/to metastasize in 22-23% of the interviews. The term ambulatory was spontaneously explained in 75% of the interviews, hepatic and metastasis/to metastasize in 22-24%. Most provided explanations were in plain language; metastasis/to metastasize and ganglion/ganglionic were among terms most frequently explained in technical language. CONCLUSION: A significant number of terms potentially unfamiliar and confusing to patients remained unclarified in training interviews conducted by senior medical students, even when they perceived the terms as technical. PRACTICE IMPLICATIONS: This exploration may offer important insights for improving future physicians' skills.

Multiple Kernel Learning of Environmental Data. Case study: Analysis and Mapping of Wind Fields

The paper presents the Multiple Kernel Learning (MKL) approach as a modelling and data exploratory tool and applies it to the problem of wind speed mapping. Support Vector Regression (SVR) is used to predict spatial variations of the mean wind speed from terrain features (slopes, terrain curvature, directional derivatives) generated at different spatial scales. Multiple Kernel Learning is applied to learn kernels for individual features and thematic feature subsets, both in the context of feature selection and optimal parameters determination. An empirical study on real-life data confirms the usefulness of MKL as a tool that enhances the interpretability of data-driven models.

Isolated disturbance of written language-acquisition as an initial symptom of epileptic aphasia in a 7-year-old child - a 3-year follow-up-study

A 7-year-old right-handed girl developed partial complex seizures with a left-sided onset. A brief period of post-ictal aphasia of the conduction type was documented before seizure control and complete normalization of oral language were obtained. We also found that she had a history of previous unexplained difficulty with written language acquisition that had occurred prior to the clinically recognized epilepsy and a subsequent loss of this ability. This rapidly improved with control of the epilepsy. The evolution of written language were been followed for 3 years, and continued improvement has occurred with fluctuations related to her epilepsy. This observation adds support to the growing body of data indicating that specific cognitive disturbances can be due to epilepsy in young children. It shows the vulnerability of skills which are in a period of active development, and the possibility that oral/written language can be differentially involved by cerebral dysfunction in the young child.

Social and individual learning of helping in humans and other species.

Helping behaviors can be innate, learned by copying others (cultural transmission) or individually learned de novo. These three possibilities are often entangled in debates on the evolution of helping in humans. Here we discuss their similarities and differences, and argue that evolutionary biologists underestimate the role of individual learning in the expression of helping behaviors in humans.