Time filter

Source Type

Porto Moniz, Portugal

Bras L.,INESC Porto | Jorge A.M.,INESC Porto | Gomes E.F.,GECAD | Duarte R.,Centro Diagnostico Pneumologico Chest Disease Center Vn Gaia
Technology and Medical Sciences, TMSi 2010 - Proceedings of the 6th International Conference on Technology and Medical Sciences | Year: 2011

We are developing a new method for the identification of rib boundaries in chest x-ray images. The identification of rib boundaries is important for radiologist diagnosis of lung diseases as TB. The radiologists use the ribs as reference for location and can be used to eliminate false positives in the detection of abnormalities. Our method automatically identifies rib boundaries from raw images through a sequence of steps using a combination of image processing techniques. Radiographs are still very relevant in practice because in Portugal and many other countries it is the first step for TB detection. We have access a large database of x-ray images provided by the pneumological screening centre (CDP) of Vila Nova de Gaia, in Portugal. © 2011 Taylor & Francis Group.

Ribeiro B.,University of Coimbra | Chen N.,GECAD
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2014

Hashing techniques have recently become the trend for accessing complex content over large data sets. With the overwhelming financial data produced today, binary embeddings are efficient tools of indexing big datasets for financial credit risk analysis. The rationale is to find a good hash function such that similar data points in Euclidean space preserve their similarities in the Hamming space for fast data retrieval. In this paper, first we use a semi-supervised hashing method to take into account the pairwise supervised information for constructing the weight adjacency graph matrix needed to learn the binarised Laplacian EigenMap. Second, we train a generalised regression neural network (GRNN) to learn the k-bits hash code. Third, the k-bits code for the test data is efficiently found in the recall phase. The results of hashing financial data show the applicability and advantages of the approach to credit risk assessment. © Springer International Publishing Switzerland 2014.

Pereira J.,UTAD | Costa C.,UTAD | Silva D.,UTAD | Varajao J.,Algoritmi | Morgado L.,GECAD
9th European Conference on eLearning 2010, ECEL 2010 | Year: 2010

While much information is available on pedagogic uses of virtual worlds, with Second Life being the most common virtual world platform in current educational literature, an organization must consider its presence in this environment as more than the mere sum of individual educational efforts. Resources need to be shared between educational stakeholders, visual navigation needs to make sense, and the sense of being within an actual organization should be conveyed (not just the sense of being within a collection of personal spaces). But there is little information on how a virtual campus for an educational organization should be structured. Virtual campi in Second Life for adult education institutions don't typically reproduce their physical counterparts. While spaces such as lecture halls, amphitheatres, meeting places, and libraries are commonly found, the specific features of the medium imply an organization of spaces and usage that differ from physical campi. For instance, navigational affordances are different (ability to fly and gravity-immune objects, for instance), as are communicational features (specific limits on the reach of voice and text communication), and user involvement (how students and teachers use the spaces). We conducted a survey of several existing Second Life campi of adult education institutions (mostly universities), to establish what spaces are present in each and how they are used and organized. In this paper, we present the overall process, and the structure and instructions for data collection by all people involved. Then we detail the various kinds of spaces (by function, not by aesthetic) found in the campi and their prevalence. We also present data on user-oriented features of the campi, and crossanalyse this with their occurrence per space and campi. This survey was part of the process for specification and development of the virtual Second Life campus for project VITA, a EC-funded project to create and experiment learning actions directed to SME' managers for development of entrepreneurship competences. Thus, we conclude with an example of how the survey results can be used to support the development of campi, by briefly presenting the campus that was developed specifically for this project.

Chen N.,GECAD | Ribeiro B.,University of Coimbra | Vieira A.S.,GECAD | Duarte J.,GECAD | Neves J.C.,University of Lisbon
ICMLC 2010 - The 2nd International Conference on Machine Learning and Computing | Year: 2010

Cost-sensitive classification algorithms that enable effective prediction, where the costs of misclassification can be very different, are crucial to creditors and auditors in credit risk analysis. Learning vector quantization (LVQ) is a powerful tool to solve bankruptcy prediction problem as a classification task. The genetic algorithm (GA) is applied widely in conjunction with artificial intelligent methods. The hybridization of genetic algorithm with existing classification algorithms is well illustrated in the field of bankruptcy prediction. In this paper, a hybrid GA and LVQ approach is proposed to minimize the expected misclassified cost under the asymmetric cost preference. Experiments on real-life French private company data show the proposed approach helps to improve the predictive performance in asymmetric cost setup. © 2010 IEEE.

Chen N.,GECAD | Ribeiro B.,University of Coimbra | Vieira A.S.,GECAD
Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) | Year: 2011

As one of the major business problems, corporate bankruptcy has been extensively studied using a large variety of statistical and machine learning approaches. However, the trajectory of bankruptcy behavior is seldom explored in the literature. In this paper, we use self-organizing map neural networks to analyze the changes of financial situation of companies in several consecutive years through a two-step clustering process. Firstly, the bankruptcy risk is characterized by a feature map, and therefore the temporal sequence is converted to the trajectory vector projected on the map. Afterwards, the trajectory map clusters the trajectory vectors to a number of evolution patterns. The approach is applied to a large database of French companies which contains the financial ratios spawning over a period of four years. Typical behaviors such as the deterioration and amelioration associated with the bankruptcy risk, as well as the influence of financial ratios can be revealed by means of visual interpretation. © 2011 Springer-Verlag.

Discover hidden collaborations