On the optimal usage of labelled examples in semi-supervised multi-class classification problems

In recent years, the performance of semi-supervised learning has been theoretically investigated. However, most of this theoretical development has focussed on binary classification problems. In this paper, we take it a step further by extending the work of Castelli and Cover [1] [2] to the multi-class paradigm. Particularly, we consider the key problem in semi-supervised learning of classifying an unseen instance x into one of K different classes, using a training dataset sampled from a mixture density distribution and composed of l labelled records and u unlabelled examples. Even under the assumption of identifiability of the mixture and having infinite unlabelled examples, labelled records are needed to determine the K decision regions. Therefore, in this paper, we first investigate the minimum number of labelled examples needed to accomplish that task. Then, we propose an optimal multi-class learning algorithm which is a generalisation of the optimal procedure proposed in the literature for binary problems. Finally, we make use of this generalisation to study the probability of error when the binary class constraint is relaxed.

Contributions to time series data mining departing from the problem of road travel time modeling

194 p.

Contributions on distance-based algorithms, multi-classifier construction and pairwise classification

179 p.

Data analysis advances in marine science for fisheries management : supervised classification application

169 p. : il. col.

Effectiveness of a Multi-Component Smoking Cessation Support Programme (McSCSP) for Patients with Severe Mental Disorders: Study Design

Only a few studies have examined the efficacy and safety of smoking cessation programmes in patients with mental disorders. The aim of this paper is to describe in detail the methodology used in the study as well as the Multi-component Smoking Cessation Support Programme in terms of pharmacological treatments and psychological interventions. An open-label 9-month follow-up study was conducted in Spain. A total of 82 clinically stable outpatients with schizophrenia, schizoaffective or bipolar disorder were enrolled. Treatment consisted of a programme specifically developed by the research team for individuals with severe mental disorders. The programme consisted of two phases: (1) weekly individual motivational therapy for 4-12 weeks, and (2) a 12-week active treatment phase. During this phase, at each study visit patients received a one- or two-week supply of medication (transdermal nicotine patches, varenicline or bupropion) with instructions on how to take it, in addition to group psychotherapy for smoking cessation. Evaluations were performed: (1) at the time of enrolment in the study, (2) during the 12-week active treatment phase of the study (weekly for the first 4 weeks and then biweekly), and (3) after the end of this phase (two follow-up assessments at weeks 12 and 24). Evaluations included: (1) smoking history, (2) substance use, (3) psychopathology, (4) adverse events, and (5) laboratory tests. The importance of this study lies in addressing a topical issue often ignored by psychiatrists: the unacceptably high rates of tobacco use in patients with severe mental disorders.