Cookies Policy
We use cookies to improve our site and your experience. By continuing to browse our site you accept our cookie policy. Find out More
Close
  • Menu
About

About

Luis Torgo is an Associate Professor of the Department of Computer Science of the Faculty of Sciences of the University of PortoPortugal. He is a senior researcher of LIAAD / INESC Tec, and a current member of the board of this research lab.

Luis Torgo is also an invited professor of the Stern Business School of the New York University where he has been collaborating in the last 3 years at the Master of Science in Business Analytics.

He has been doing research in the area of Data Mining and Machine Learning since 1990, and has published over 100 papers in several foruns of these areas. Luis Torgo is the author of the widely acclaimed Data Mining with R book published by CRC Press in 2010 with a strongly revised second edition published in January of 2017. He has been involved in many research projects under different roles and involving different types of organizations.

His current broad research interests revolve around analyzing data from dynamic environments, with a particular focus on time and space-time dependent data sets, in the search for unexpected events. In terms of application domains his research is frequently linked with ecological/biological as well as financial domains.

Luis Torgo main contributions to the state of the art on data mining and machine learning are related with tree-based regression methods and more recently with utility-based forecasting methods.

He has a strong experience of teaching different subjects at different academic levels but also in non-academic settings. He is frequently invited for giving short courses on using R for data mining around the world.

Luis Torgo is the CEO and one of the founding partners of KNOYDA a company devoted to training and consulting within data science.

Interest
Topics
Details

Details

  • Name

    Luís Torgo
  • Cluster

    Computer Science
  • Role

    Senior Researcher
  • Since

    01st January 2008
007
Publications

2019

On Feature Selection and Evaluation of Transportation Mode Prediction Strategies

Authors
Etemad, M; Soares, A; Matwin, S; Torgo, L;

Publication
Proceedings of the Workshops of the EDBT/ICDT 2019 Joint Conference, EDBT/ICDT 2019, Lisbon, Portugal, March 26, 2019.

Abstract

2019

Pre-processing approaches for imbalanced distributions in regression

Authors
Branco, P; Torgo, L; Ribeiro, RP;

Publication
NEUROCOMPUTING

Abstract
Imbalanced domains are an important problem frequently arising in real world predictive analytics. A significant body of research has addressed imbalanced distributions in classification tasks, where the target variable is nominal. In the context of regression tasks, where the target variable is continuous, imbalanced distributions of the target variable also raise several challenges to learning algorithms. Imbalanced domains are characterized by: (1) a higher relevance being assigned to the performance on a subset of the target variable values; and (2) these most relevant values being underrepresented on the available data set. Recently, some proposals were made to address the problem of imbalanced distributions in regression. Still, this remains a scarcely explored issue with few existing solutions. This paper describes three new approaches for tackling the problem of imbalanced distributions in regression tasks. We propose the adaptation to regression tasks of random over-sampling and introduction of Gaussian Noise, and we present a new method called WEighted Relevance-based Combination Strategy (WERCS). An extensive set of experiments provides empirical evidence of the advantage of using the proposed strategies and, in particular, the WERCS method. We analyze the impact of different data characteristics in the performance of the methods. A data repository with 15 imbalanced regression data sets is also provided to the research community.

2019

Arbitrage of forecasting experts

Authors
Cerqueira, V; Torgo, L; Pinto, F; Soares, C;

Publication
Machine Learning

Abstract
Forecasting is an important task across several domains. Its generalised interest is related to the uncertainty and complex evolving structure of time series. Forecasting methods are typically designed to cope with temporal dependencies among observations, but it is widely accepted that none is universally applicable. Therefore, a common solution to these tasks is to combine the opinion of a diverse set of forecasts. In this paper we present an approach based on arbitrating, in which several forecasting models are dynamically combined to obtain predictions. Arbitrating is a metalearning approach that combines the output of experts according to predictions of the loss that they will incur. We present an approach for retrieving out-of-bag predictions that significantly improves its data efficiency. Finally, since diversity is a fundamental component in ensemble methods, we propose a method for explicitly handling the inter-dependence between experts when aggregating their predictions. Results from extensive empirical experiments provide evidence of the method’s competitiveness relative to state of the art approaches. The proposed method is publicly available in a software package. © 2018, The Author(s).

2018

Potential of dissimilatory nitrate reduction pathways in polycyclic aromatic hydrocarbon degradation

Authors
Ribeiro, H; de Sousa, T; Santos, JP; Sousa, AGG; Teixeira, C; Monteiro, MR; Salgado, P; Mucha, AP; Almeida, CMR; Torgo, L; Magalhaes, C;

Publication
Chemosphere

Abstract
This study investigates the potential of an indigenous estuarine microbial consortium to degrade two polycyclic aromatic hydrocarbons (PAHs), naphthalene and fluoranthene, under nitrate-reducing conditions. Two physicochemically diverse sediment samples from the Lima Estuary (Portugal) were spiked individually with 25 mg L-1 of each PAH in laboratory designed microcosms. Sediments without PAHs and autoclaved sediments spiked with PAHs were run in parallel. Destructive sampling at the beginning and after 3, 6, 12, 30 and 63 weeks incubation was performed. Naphthalene and fluoranthene levels decreased over time with distinct degradation dynamics varying with sediment type. Next-generation sequencing (NGS) of 16 S rRNA gene amplicons revealed that the sediment type and incubation time were the main drivers influencing the microbial community structure rather than the impact of PAH amendments. Predicted microbial functional analyses revealed clear shifts and interrelationships between genes involved in anaerobic and aerobic degradation of PAHs and in the dissimilatory nitrate-reducing pathways (denitrification and dissimilatory nitrate reduction to ammonium - DNRA). These findings reinforced by clear biogeochemical denitrification signals (NO3 - consumption, and NH4 + increased during the incubation period), suggest that naphthalene and fluoranthene degradation may be coupled with denitrification and DNRA metabolism. The results of this study contribute to the understanding of the dissimilatory nitrate-reducing pathways and help uncover their involvement in degradation of PAHs, which will be crucial for directing remediation strategies of PAH-contaminated anoxic sediments. © 2018 Elsevier Ltd

2018

How to evaluate sentiment classifiers for Twitter time-ordered data?

Authors
Mozetic, I; Torgo, L; Cerqueira, V; Smailovic, J;

Publication
PLoS ONE

Abstract
Social media are becoming an increasingly important source of information about the public mood regarding issues such as elections, Brexit, stock market, etc. In this paper we focus on sentiment classification of Twitter data. Construction of sentiment classifiers is a standard text mining task, but here we address the question of how to properly evaluate them as there is no settled way to do so. Sentiment classes are ordered and unbalanced, and Twitter produces a stream of time-ordered data. The problem we address concerns the procedures used to obtain reliable estimates of performance measures, and whether the temporal ordering of the training and test data matters. We collected a large set of 1.5 million tweets in 13 European languages. We created 138 sentiment models and out-of-sample datasets, which are used as a gold standard for evaluations. The corresponding 138 in-sample data-sets are used to empirically compare six different estimation procedures: three variants of cross-validation, and three variants of sequential validation (where test set always follows the training set). We find no significant difference between the best cross-validation and sequential validation. However, we observe that all cross-validation variants tend to overestimate the performance, while the sequential methods tend to underestimate it. Standard cross-validation with random selection of examples is significantly worse than the blocked cross-validation, and should not be used to evaluate classifiers in time-ordered data scenarios. © 2018 Mozetic et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.

Supervised
thesis

2017

Prediction and Ranking of Highly Popular Web Content

Author
Nuno Miguel Pereira Moniz

Institution
IES_Outra

2016

Utility-based Predictive analytics

Author
Paula Alexandra de Oliveira Branco

Institution
UP-FCUP

2016

Domain Oriented Biclustering Validation

Author
Carlos Alberto Magalhães Leite

Institution
UP-FCUP

2016

Lexicon Expansion System for Domain and Time Oriented Sentiment Analysis

Author
Nuno Ricardo Pinheiro da Silva Guimarães

Institution
UP-FCUP

2015

Actionable Forecasting and Activity Monitoring: Applications to Financial Trading

Author
Luís Carlos Gouveia Baía

Institution
UP-FCUP