Securing ad hoc networks. mechanisms and techniques
Tesis doctoral de Helena RifÁ Pous This thesis presents a study about security mechanisms and techniques to develop reliable and […]
Accede a todas las tesis doctorales sobre Politécnica de catalunya mas relevantes.
Tesis doctoral de Helena RifÁ Pous This thesis presents a study about security mechanisms and techniques to develop reliable and […]
Tesis doctoral de Carlos Jorge Henriques Ferreira The research is a contribute to the knowledge about the housing processes in
Tesis doctoral de Isabel Victoria Martín Faus To achieve the provision of end-to-end quality of service (qos) for multimedia services,
Tesis doctoral de Delfina Muñoz Cervantes The cost of high efficiency crystalline silicon (c-si) modules is hindering the progress of
Tesis doctoral de Joaquin Otón Pérez In this dissertation, we have addressed a range of motivated research objectives related to
Tesis doctoral de Rodolfo Rodríguez Baracaldo Un tema de gran interés científico y tecnológico es la obtención de materiales con
Tesis doctoral de Víctor Eusebi Muñoz Cinca In this thesis we analyze the whole terminal operating processes being conducted in
Tesis doctoral de Montserrat Padró Cirera This thesis is focused on the study and use of causal state splitting reconstruction (cssr) algorithm for natural language processing (nlp) tasks. cssr is an algorithm that captures patterns from data building automata in the form of visible markov models. it is based on the principles of computational mechanics and takes advantage of many properties of causal state theory. one of the main advantages of cssr with respect to markov models is that it builds states containing more than one $n$Âgram (called history in computational mechanics), so the obtained automata are much smaller than the equivalent markov model. in this work, we first study the behavior of the algorithm when learning patterns related to nlp tasks but without performing any annotation task. this first experiments are useful to understand the parameters that affect the algorithm and to check that it is able to capture the patterns present in natural language sentences. secondly, we propose a way to apply cssr to nlp annotation tasks. the algorithm is not originally conceived to use the hidden information necessary for annotation tasks, so we devised a way to introduce it into the system in order to obtain automata including this information that can be afterwards used to annotate new text. also, some methods to deal with unseen events and a modification of the algorithm to make it more suitable for nlp tasks have been presented and tested. these three aspects conform the first line of contributions of this research, altogether with aÂ
Tesis doctoral de Fausto José Canhoto De Pavia Vieira Wireless communications have always been characterized by the challenge of fully
Tesis doctoral de Juan Felipe Núñez Espinoza The concept of rural social networks, as a social relational structure made of