Abstract
Video segmentation is one of the most important parts of a vision system which allows partitioning each frame into homogeneous regions that share a common property. This work proposes a new methodology that aggregates three different techniques: background subtraction, region growing and a pseudo Fuzzy colour model to define colour subspaces that characterize each class. In addition, the pseudo Fuzzy colour model allows a given colour to belong to more than one class and enables the expansion ofthe classes through a dynamic model based on belonging and persistence information. In case of shared colours among classes, regional features are searched in order to determine the object's class. Tests with test and real videos of sports footages show promising results.

Abstract

Abstract

Abstract
This paper envisages showing the potential of innovative non-invasive techniques based on affordable and easily operated instrumentation as well as user-friendly computer aided algorithms in the screening of cardiovascular (CV) diseases. These techniques are based on the assumption that arterial stiffness is currently an important predicator of the CV diseases development and can be assessed by analyzing the arterial pressure waveform (APW). A previously developed PZ based device for non-invasive APW capture is currently under test in clinical environment, using a heterogeneous population constituted by healthy and unhealthy subjects. A dedicated Matlab analysis tool was designed and developed to extract relevant information and further APW analysis. Several recordings of the APW in the same day and in consecutive months are being performed by trained observers, to evaluate its reproducibility. Data mining analysis is subsequently the last task where the Weka 3-6-5 package software is used. The usefulness of developing data mining algorithms for cardiovascular applications can benefit the CV screenings contributing for the early identification of arterial stiffness related patterns.

Abstract
The basic objective of Transmission Expansion Planning (TEP) is to schedule a number of transmission projects along an extended planning horizon minimizing the network construction and operational costs while satisfying the requirement of delivering power safely and reliably to load centres along the horizon. This principle is quite simple, but the complexity of the problem and the impact on society transforms TEP on a challenging issue. This paper describes a new approach to solve the dynamic TEP problem, based on an improved discrete integer version of the Evolutionary Particle Swarm Optimization (EPSO) meta-heuristic algorithm. The paper includes sections describing in detail the EPSO enhanced approach, the mathematical formulation of the TEP problem, including the objective function and the constraints, and a section devoted to the application of the developed approach to this problem. Finally, the use of the developed approach is illustrated using a case study based on the IEEE 24 bus 38 branch test system.

Abstract
When searching for patterns on data streams, we come across perennial (dynamic) objects that evolve over time. These objects are encountered repeatedly and each time with different definition and values. Examples are (a) companies registered at stock exchange and reporting their progress at the end of each year, and (b) students whose performance is evaluated at the end of each semester. On such data, domain experts also pose questions on how the individual objects will evolve: would it be beneficial to invest in a given company, given both the company's individual performance thus far and the drift experienced in the model? Or, how will a given student perform next year, given the performance variations observed thus far? While there is much research on how models evolve/change over time [Ntoutsi et al., 2011a], little is done to predict the change of individual objects when the states are not known a priori. In this work, we propose a framework that learns the clusters to which the objects belong at each moment, uses them as ad hoc states in a state-transition graph, and then learns a mixture model of Markov Chains, which predicts the next most likely state/cluster per object. We report on our evaluation on synthetic and real datasets. © Springer-Verlag Berlin Heidelberg 2012.