Optimización y aproximación al problema de la regresión simbólica a través de Straight Line Programs (SLP's) y algoritmos genéticos : Entrenamiento genético de SLP's

El problema de la regresión simbólica consiste en el aprendizaje, a partir de un conjunto muestra de datos obtenidos experimentalmente, de una función desconocida. Los métodos evolutivos han demostrado su eficiencia en la resolución de instancias de dicho problema. En este proyecto se propone una nueva estrategia evolutiva, a través de algoritmos genéticos, basada en una nueva estructura de datos denominada Straight Line Program (SLP) y que representa en este caso expresiones simbólicas. A partir de un SLP universal, que depende de una serie de parámetros cuya especialización proporciona SLP's concretos del espacio de búsqueda, la estrategia trata de encontrar los parámetros óptimos para que el SLP universal represente la función que mejor se aproxime al conjunto de puntos muestra. De manera conceptual, este proyecto consiste en un entrenamiento genético del SLP universal, utilizando los puntos muestra como conjunto de entrenamiento, para resolver el problema de la regresión simbólica.

Algoritmo genético em química

Genetic algorithm is an optimization technique based on Darwin evolution theory. In last years its application in chemistry is increasing significantly due the special characteristics for optimization of complex systems. The basic principles and some further modifications implemented to improve its performance are presented, as well as a historical development. A numerical example of a function optimization is also shown to demonstrate how the algorithm works in an optimization process. Finally several chemistry applications realized until now is commented to serve as parameter to future applications in this field.

Detection of severe obstructive sleep apnea through voice analysis

tThis paper deals with the potential and limitations of using voice and speech processing to detect Obstruc-tive Sleep Apnea (OSA). An extensive body of voice features has been extracted from patients whopresent various degrees of OSA as well as healthy controls. We analyse the utility of a reduced set offeatures for detecting OSA. We apply various feature selection and reduction schemes (statistical rank-ing, Genetic Algorithms, PCA, LDA) and compare various classifiers (Bayesian Classifiers, kNN, SupportVector Machines, neural networks, Adaboost). S-fold crossvalidation performed on 248 subjects showsthat in the extreme cases (that is, 127 controls and 121 patients with severe OSA) voice alone is able todiscriminate quite well between the presence and absence of OSA. However, this is not the case withmild OSA and healthy snoring patients where voice seems to play a secondary role. We found that thebest classification schemes are achieved using a Genetic Algorithm for feature selection/reduction.

Statistical Optimum Design of Heat Exchangers

The optimal design of a heat exchanger system is based on given model parameters together with given standard ranges for machine design variables. The goals set for minimizing the Life Cycle Cost (LCC) function which represents the price of the saved energy, for maximizing the momentary heat recovery output with given constraints satisfied and taking into account the uncertainty in the models were successfully done. Nondominated Sorting Genetic Algorithm II (NSGA-II) for the design optimization of a system is presented and implemented inMatlab environment. Markov ChainMonte Carlo (MCMC) methods are also used to take into account the uncertainty in themodels. Results show that the price of saved energy can be optimized. A wet heat exchanger is found to be more efficient and beneficial than a dry heat exchanger even though its construction is expensive (160 EUR/m2) compared to the construction of a dry heat exchanger (50 EUR/m2). It has been found that the longer lifetime weights higher CAPEX and lower OPEX and vice versa, and the effect of the uncertainty in the models has been identified in a simplified case of minimizing the area of a dry heat exchanger.

Aplicação de algoritmos genéticos na seleção de variáveis em espectroscopia no infravermelho médio: determinação simultânea de glicose, maltose e frutose

Genetic algorithm was used for variable selection in simultaneous determination of mixtures of glucose, maltose and fructose by mid infrared spectroscopy. Different models, using partial least squares (PLS) and multiple linear regression (MLR) with and without data pre-processing, were used. Based on the results obtained, it was verified that a simpler model (multiple linear regression with variable selection by genetic algorithm) produces results comparable to more complex methods (partial least squares). The relative errors obtained for the best model was around 3% for the sugar determination, which is acceptable for this kind of determination.

Seleção de variáveis em QSAR

The process of building mathematical models in quantitative structure-activity relationship (QSAR) studies is generally limited by the size of the dataset used to select variables from. For huge datasets, the task of selecting a given number of variables that produces the best linear model can be enormous, if not unfeasible. In this case, some methods can be used to separate good parameter combinations from the bad ones. In this paper three methodologies are analyzed: systematic search, genetic algorithm and chemometric methods. These methods have been exposed and discussed through practical examples.

Estudo do estado fundamental de aglomerados de silício via redes neurais

We introduce a global optimization method based on the cooperation between an Artificial Neural Net (ANN) and Genetic Algorithm (GA). We have used ANN to select the initial population for the GA. We have tested the new method to predict the ground-state geometry of silicon clusters. We have described the clusters as a piling of plane structures. We have trained three ANN architectures and compared their results with those of pure GA. ANN strongly reduces the total computational time. For Si10, it gained a factor of 5 in search speed. This method can be easily extended to other optimization problems.

Development a distributed simulation environment on a cluster of workstations

Simulation has traditionally been used for analyzing the behavior of complex real world problems. Even though only some features of the problems are considered, simulation time tends to become quite high even for common simulation problems. Parallel and distributed simulation is a viable technique for accelerating the simulations. The success of parallel simulation depends heavily on the combination of the simulation application, algorithm and message population in the simulation is sufficient, no additional delay is caused by this environment. In this thesis a conservative, parallel simulation algorithm is applied to the simulation of a cellular network application in a distributed workstation environment. This thesis presents a distributed simulation environment, Diworse, which is based on the use of networked workstations. The distributed environment is considered especially hard for conservative simulation algorithms due to the high cost of communication. In this thesis, however, the distributed environment is shown to be a viable alternative if the amount of communication is kept reasonable. Novel ideas of multiple message simulation and channel reduction enable efficient use of this environment for the simulation of a cellular network application. The distribution of the simulation is based on a modification of the well known Chandy-Misra deadlock avoidance algorithm with null messages. The basic Chandy Misra algorithm is modified by using the null message cancellation and multiple message simulation techniques. The modifications reduce the amount of null messages and the time required for their execution, thus reducing the simulation time required. The null message cancellation technique reduces the processing time of null messages as the arriving null message cancels other non processed null messages. The multiple message simulation forms groups of messages as it simulates several messages before it releases the new created messages. If the message population in the simulation is suffiecient, no additional delay is caused by this operation A new technique for considering the simulation application is also presented. The performance is improved by establishing a neighborhood for the simulation elements. The neighborhood concept is based on a channel reduction technique, where the properties of the application exclusively determine which connections are necessary when a certain accuracy for simulation results is required. Distributed simulation is also analyzed in order to find out the effect of the different elements in the implemented simulation environment. This analysis is performed by using critical path analysis. Critical path analysis allows determination of a lower bound for the simulation time. In this thesis critical times are computed for sequential and parallel traces. The analysis based on sequential traces reveals the parallel properties of the application whereas the analysis based on parallel traces reveals the properties of the environment and the distribution.

Niche Differentiation and Fine-Scale Projections for Argentine Ants Based on Remotely Sensed Data

Modeling ecological niches of species is a promising approach for predicting the geographic potential of invasive species in new environments. Argentine ants (Linepithema humile) rank among the most successful invasive species: native to South America, they have invaded broad areas worldwide. Despite their widespread success, little is known about what makes an area susceptible - or not - to invasion. Here, we use a genetic algorithm approach to ecological niche modeling based on high-resolution remote-sensing data to examine the roles of niche similarity and difference in predicting invasions by this species. Our comparisons support a picture of general conservatism of the species' ecological characteristics, in spite of distinct geographic and community contexts

Lautanauhaohjeen muodostus tavoitekuviosta geneettisellä algoritmilla

Lautanauhatekniikka on monipuolinen menetelmä esimerkiksi kuvioitujen nauhojen kutomiseen, mutta uusien kuvioaiheiden suunnittelu, tai aloittelijalle jo valmiiden ohjeettomien kuviomallien jäljittely, voi helposti käydä työlääksi menetelmän ominaispiirteiden johdosta. Tämän työn tavoitteena oli kehittää ohjelmallinen työkalu auttamaan näissä ongelmissa automatisoimalla kudontaohjeen etsintä käyttäjän laatimalle tavoitekuviolle. Ratkaisumenetelmän perustaksi valittiin geneettinen algoritmi, minkä johdosta työn keskeisintutkimusongelma oli kartoittaa algoritmin perusoperaatioiden parametrien ja tavoitekuvion kompleksisuuden keskinäisiä riippuvuuksia riittävästi toimivien arvosuositusten antamiseen ohjelman tulevassa käytännön käytössä. Työssä ei kehitetty sovellusalueeseen mukautettuja evoluutiooperaatioita, vaan keskityttiin luomaan hyvin tunnetuista elementeistä perusta, jota voi myöhemmin kehittää eteenpäin.

Quantitative structure-retention relationships analysis of retention index of essential oils

Genetic algorithm and multiple linear regression (GA-MLR), partial least square (GA-PLS), kernel PLS (GA-KPLS) and Levenberg-Marquardt artificial neural network (L-M ANN) techniques were used to investigate the correlation between retention index (RI) and descriptors for 116 diverse compounds in essential oils of six Stachys species. The correlation coefficient LGO-CV (Q²) between experimental and predicted RI for test set by GA-MLR, GA-PLS, GA-KPLS and L-M ANN was 0.886, 0.912, 0.937 and 0.964, respectively. This is the first research on the QSRR of the essential oil compounds against the RI using the GA-KPLS and L-M ANN.

Application of GA-PLS and GA-KPLS calculations for the prediction of the retention indices of essential oils

Genetic algorithm and partial least square (GA-PLS) and kernel PLS (GA-KPLS) techniques were used to investigate the correlation between retention indices (RI) and descriptors for 117 diverse compounds in essential oils from 5 Pimpinella species gathered from central Turkey which were obtained by gas chromatography and gas chromatography-mass spectrometry. The square correlation coefficient leave-group-out cross validation (LGO-CV) (Q²) between experimental and predicted RI for training set by GA-PLS and GA-KPLS was 0.940 and 0.963, respectively. This indicates that GA-KPLS can be used as an alternative modeling tool for quantitative structure-retention relationship (QSRR) studies.

Testing Platform Development for Large Scale Solar Integration

Currently, a high penetration level of Distributed Generations (DGs) has been observed in the Danish distribution systems, and even more DGs are foreseen to be present in the upcoming years. How to utilize them for maintaining the security of the power supply under the emergency situations, has been of great interest for study. This master project is intended to develop a control architecture for studying purposes of distribution systems with large scale integration of solar power. As part of the EcoGrid EU Smart Grid project, it focuses on the system modelling and simulation of a Danish representative LV network located in Bornholm island. Regarding the control architecture, two types of reactive control techniques are implemented and compare. In addition, a network voltage control based on a tap changer transformer is tested. The optimized results after applying a genetic algorithm to five typical Danish domestic loads are lower power losses and voltage deviation using Q(U) control, specially with large consumptions. Finally, a communication and information exchange system is developed with the objective of regulating the reactive power and thereby, the network voltage remotely and real-time. Validation test of the simulated parameters are performed as well.

Non-Linear Unsteady Aerodynamic Response Approximation Using Multi-Layer Functionals

Non-linear functional representation of the aerodynamic response provides a convenient mathematical model for motion-induced unsteady transonic aerodynamic loads response, that accounts for both complex non-linearities and time-history effects. A recent development, based on functional approximation theory, has established a novel functional form; namely, the multi-layer functional. For a large class of non-linear dynamic systems, such multi-layer functional representations can be realised via finite impulse response (FIR) neural networks. Identification of an appropriate FIR neural network model is facilitated by means of a supervised training process in which a limited sample of system input-output data sets is presented to the temporal neural network. The present work describes a procedure for the systematic identification of parameterised neural network models of motion-induced unsteady transonic aerodynamic loads response. The training process is based on a conventional genetic algorithm to optimise the network architecture, combined with a simplified random search algorithm to update weight and bias values. Application of the scheme to representative transonic aerodynamic loads response data for a bidimensional airfoil executing finite-amplitude motion in transonic flow is used to demonstrate the feasibility of the approach. The approach is shown to furnish a satisfactory generalisation property to different motion histories over a range of Mach numbers in the transonic regime.

Modelagem da distribuição geográfica de espécies lenhosas de cerrado no Estado de São Paulo

Atualmente vêm sendo desenvolvidas e utilizadas várias técnicas de modelagem de distribuição geográfica de espécies com os mais variados objetivos. Algumas dessas técnicas envolvem modelagem baseada em análise ambiental, nas quais os algoritmos procuram por condições ambientais semelhantes àquelas onde as espécies foram encontradas, resultando em áreas potenciais onde as condições ambientais seriam propícias ao desenvolvimento dessas espécies. O presente estudo trata do uso da modelagem preditiva de distribuição geográfica de espécies nativas, através da utilização de algoritmo genético, como ferramenta para auxiliar o entendimento dos padrões de distribuição do bioma cerrado no Estado de São Paulo. A metodologia empregada e os resultados obtidos foram considerados satisfatórios para a geração de modelos de distribuição geográfica de espécies vegetais, baseados em dados abióticos, para as regiões de estudo. A eficácia do modelo em predizer a ocorrência de espécies do cerrado é maior se forem utilizados apenas pontos de amostragem com fisionomias de cerrado, excluindo-se áreas de transição. Para minimizar problemas decorrentes da falta de convergência do algoritmo utilizado GARP ("Genetic Algorithm for Rule Set Production"), foram gerados 100 modelos para cada espécie modelada. O uso de modelagem pode auxiliar no entendimento dos padrões de distribuição de um bioma ou ecossistema em uma análise regional.