Don Bosco Fernandez

Fernandez Lab - Publications

Support Vector Machines Lecturer : Arturo Fernandez Scribe : Arturo Fernandez 1 Support Vector

Consider a binary classification prediction problem. Training data are given and we denote them observations as T = {X,y} = {xi, yi}i=1, where xi ∈ R, yi ∈ {+1,−1} (accordingly X ∈ Rm×n). First, we will consider the case where the two classes are linearly separable. That is, by an n-dimensional decision boundary which is the result of an n + 1-dimensional hyperplane). Furthermore, since there c...

05. Fernandez Frias

Introduction: performing anal endosonography in complex fistula-in-ano allows us to design a personalized surgical strategy in each case, thereby improving results. However, there are doubts in the literature as to its utility in recurrent complex fistulas. The aim of this study was to compare the utility of anal ultrasonography in the study of primary versus recurrent complex fistula-inano. Pa...

146-155 FERNANDEZ MONTALVO.indd

This study explores the prevalence of criminal behaviour in patients addicted to drugs who are in treatment. A sample of 252 addicted patients (203 male and 49 female) who sought outpatient treatment at a specialized centre was assessed. Information on criminal behaviours, socio-demographic factors, consumption factors (assessed by the EuropAsi), psychopathological factors (assessed by SCL-90-R...

Fernandez-Soria 1_8

Invasiveness and metastatic potential are the two most important properties defining malignancy. The adenovirus E1A (Ad-E1A) gene has a dual effect as a proliferative gene and as a tumor-suppressor gene, decreasing tumor growth and the metastatic potential of malignant cells. In order to study genes related with the antimetastatic effect of Ad-E1A in human cells, we performed a microarray analy...