Incorporating label dependency into the binary relevance framework for multi-label classification

In multi-label classification, examples can be associated with multiple labels simultaneously. The task of learning from multi-label data can be addressed by methods that transform the multi-label classification problem into several single-label classification problems. The binary relevance approach is one of these methods, where the multi-label learning task is decomposed into several independent binary classification problems, one for each label in the set of labels, and the final labels for each example are determined by aggregating the predictions from all binary classifiers. However, this approach fails to consider any dependency among the labels. Aiming to accurately predict label combinations, in this paper we propose a simple approach that enables the binary classifiers to discover existing label dependency by themselves. An experimental study using decision trees, a kernel method as well as Naive Bayes as base-learning techniques shows the potential of the proposed approach to improve the multi-label classification performance.

Computational tools for comparing asymmetric GARCH models via Bayes factors

In this paper we use Markov chain Monte Carlo (MCMC) methods in order to estimate and compare GARCH models from a Bayesian perspective. We allow for possibly heavy tailed and asymmetric distributions in the error term. We use a general method proposed in the literature to introduce skewness into a continuous unimodal and symmetric distribution. For each model we compute an approximation to the marginal likelihood, based on the MCMC output. From these approximations we compute Bayes factors and posterior model probabilities. (C) 2012 IMACS. Published by Elsevier B.V. All rights reserved.

Plasma cortisol in first episode drug-naive mania: Differential levels in euphoric versus irritable mood

Background: Dysregulation of HPA axis has been widely described in subjects with bipolar disorder (BD), including changes in cortisol levels during mood episodes and euthymia. However, most of the studies were done with medicated BD patients with variable length of illness, which was shown to interfere on peripheral cortisol levels. Therefore, the present study aims to evaluate plasma cortisol levels in drug-naive BD subjects during the first manic episode, as well as investigate the relationship between plasma cortisol levels and manic symptomatology. Methods: Twenty-six drug-naive patients were enrolled meeting criteria for a first manic episode in bipolar I disorder. Severity of mania was assessed using the Young Mania Rating Scale (YMRS). The control group included 27 healthy subjects matched by age and gender. Cortisol was quantified using a direct radioimmunoassay. Results: Plasma cortisol levels were decreased during first manic episode compared to healthy controls. Higher cortisol levels were positively associated with the presence of irritability (dysphoria), while elated mania showed lower cortisol levels compared to controls. Limitation: Data including larger samples are lacking. Conclusion: Higher cortisol in dysphoric mania compared to predominantly elated/euphoric mania may indicate a clinical and neurobiological polymorphic phenomenon, potentially involving a higher biological sensitivity to stress in the presence of irritable mood. The present findings highlight the importance to add a dimensional approach to the traditional categorical diagnosis for future neurobiological studies in BD. (C) 2011 Elsevier B.V. All rights reserved.

Obsessive-compulsive symptom dimensions correlate to specific gray matter volumes in treatment-naive patients

Background: Clinical and sociodemographic findings have supported that OCD is heterogeneous and composed of multiple potentially overlapping and stable symptom dimensions. Previous neuroimaging investigations have correlated different patterns of OCD dimension scores and gray matter (GM) volumes. Despite their relevant contribution, some methodological limitations, such as patient's previous medication intake, may have contributed to inconsistent findings. Method: Voxel-based morphometry was used to investigate correlations between regional GM volumes and symptom dimensions severity scores in a sample of 38 treatment-naive OCD patients. Several standardized instruments were applied, including an interview exclusively developed for assessing symptom dimensions severity (DY-BOCS). Results: Scores on the "aggression" dimension were positively correlated with GM volumes in lateral parietal cortex in both hemispheres and negatively correlated with bilateral insula, left putamen and left inferior OFC. Scores on the "sexual/religious" dimension were positively correlated with GM volumes within the right middle lateral OFC and right DLPFC and negatively correlated with bilateral ACC. Scores on the "hoarding" dimension were positively correlated with GM volumes in the left superior lateral OFC and negatively correlated in the right parahippocampal gyrus. No significant correlations between GM volumes and the "contamination" or "symmetry" dimensions were found. Conclusions: Building upon preexisting findings, our data with treatment-naive OCD patients have demonstrated distinct GM substrates implicated in both cognitive and emotion processing across different OCS dimensions. (C) 2012 Elsevier Ltd. All rights reserved.