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

I got a BSc in Physics before I went on to get a MSc in Computer Science at the Faculty of Sciences, University of Porto. I am now working towards a PhD in Informatics through the MAP-I program.

My research interests are mostly in Data Mining / Machine Learning, focusing on predictive analytics for dependent data.

Interest
Topics
Details

Details

  • Name

    Mariana Rafaela Oliveira
  • Cluster

    Computer Science
  • Role

    Research Assistant
  • Since

    01st February 2014
Publications

2017

Dynamic and Heterogeneous Ensembles for Time Series Forecasting

Authors
Cerqueira, V; Torgo, L; Oliveira, M; Pfahringer, B;

Publication
2017 IEEE International Conference on Data Science and Advanced Analytics, DSAA 2017, Tokyo, Japan, October 19-21, 2017

Abstract

2016

Predicting Wildfires Propositional and Relational Spatio-Temporal Pre-processing Approaches

Authors
Oliveira, M; Torgo, L; Costa, VS;

Publication
DISCOVERY SCIENCE, (DS 2016)

Abstract
We present and evaluate two different methods for building spatio-temporal features: a propositional method and a method based on propositionalisation of relational clauses. Our motivating application, a regression problem, requires the prediction of the fraction of each Portuguese parish burnt yearly by wildfires - a problem with a strong socio-economic and environmental impact in the country. We evaluate and compare how these methods perform individually and combined together. We successfully use under-sampling to deal with the high skew in the data set. We find that combining the approaches significantly improves the similar results obtained by each method individually.

2016

Development of an autonomous system for integrated marine monitoring

Authors
Catarina, M; Ana Paula, M; Maria, C; Hugo, R; Cristina, A; Isabel, A; Sandra, R; Teresa, B; Sérgio, L; Antonina, DS; Alexandra, S; Cátia, B; Sónia, C; Raquel, M; Catarina, C; André, D; Hugo, F; Ireneu, D; Luís, T; Mariana, O; Nuno, D; Pedro, J; Alfredo, M; Eduardo, S;

Publication
Frontiers in Marine Science

Abstract

2015

Automatized and desktop AC-susceptometer for the in situ and real time monitoring of magnetic nanoparticles' synthesis by coprecipitation

Authors
Fernandez Garcia, MP; Teixeira, JM; Machado, P; Oliveira, MRFF; Maia, JM; Pereira, C; Pereira, AM; Freire, C; Araujo, JP;

Publication
REVIEW OF SCIENTIFIC INSTRUMENTS

Abstract
The main purpose of this work was to design, develop, and construct a simple desktop AC susceptometer to monitor in situ and in real time the coprecipitation synthesis of magnetic nanoparticles. The design incorporates one pair of identical pick-up sensing coils and one pair of Helmholtz coils. The picked up signal is detected by a lock-in SR850 amplifier that measures the in-and out-of-phase signals. The apparatus also includes a stirrer with 45 degrees-angle blades to promote the fast homogenization of the reaction mixture. Our susceptometer has been successfully used to monitor the coprecipitation reaction for the synthesis of iron oxide nanoparticles. (C) 2015 AIP Publishing LLC.

2014

Ensembles for Time Series Forecasting

Authors
Oliveira, M; Torgo, L;

Publication
Proceedings of the Sixth Asian Conference on Machine Learning, ACML 2014, Nha Trang City, Vietnam, November 26-28, 2014.

Abstract
This paper describes a new type of ensembles that aims at improving the predictive performance of these approaches in time series forecasting. Ensembles are recognised as one of the most successful approaches to prediction tasks. Previous theoretical studies of ensembles have shown that one of the key reasons for this performance is diversity among ensemble members. Several methods exist to generate diversity. The key idea of the work we are presenting here is to propose a new form of diversity generation that explores some specific properties of time series prediction tasks. Our hypothesis is that the resulting ensemble members will be better at addressing different dynamic regimes of time series data. Our large set of experiments confirms that the methods we have explored for generating diversity are able to improve the performance of the equivalent ensembles with standard diversity generation procedures. © 2014 M. Oliveira & L. Torgo.