Joaquim Fernando Pinto da Costa's Annual Report
Year:
2017
Brief description of the research activities:
Clustering of longitudinal data, Unimodal discriminant analysis and logistic regression, Spline-based methods for classification of very high dimensional datasets, Supervised Classication Applied to Hot-Spot Detection in Protein-Protein Interfaces. Combining ranking with traditional methods for ordinal class imbalance. Tackling Class Imbalance with Ranking.
Talks / Seminars / Courses :
Communications in national conferences
Title:
Alguns desenvolvimentos em Clustering de Trajectórias Longitudinais
Name of the event:
SPE 2017
Talk:
Contributed talk
Date:
18.10.2017 to 22.10.2017
Host institution:
ISCTE
Country:
Portugal
Location / City:
Lisbon
Title:
A Model-based Approach for the Clustering of Gene-Expression Data
Name of the event:
Symposium on Big Data
Talk:
Contributed talk
Date:
02.11.2017 to 04.11.2017
Country:
Portugal
Location / City:
Lisbon
Communications in international conferences
Work visits:
Sabbatical visit to Penn State University (USA), department of Statistics.