Machine learning Gaussian short rate

Dissertação para obtenção do Grau de Doutor em Estatística e Gestão do Risco

Constraint reasoning with local search for continuous optimization

Optimization is a very important field for getting the best possible value for the optimization function. Continuous optimization is optimization over real intervals. There are many global and local search techniques. Global search techniques try to get the global optima of the optimization problem. However, local search techniques are used more since they try to find a local minimal solution within an area of the search space. In Continuous Constraint Satisfaction Problems (CCSP)s, constraints are viewed as relations between variables, and the computations are supported by interval analysis. The continuous constraint programming framework provides branch-and-prune algorithms for covering sets of solutions for the constraints with sets of interval boxes which are the Cartesian product of intervals. These algorithms begin with an initial crude cover of the feasible space (the Cartesian product of the initial variable domains) which is recursively refined by interleaving pruning and branching steps until a stopping criterion is satisfied. In this work, we try to find a convenient way to use the advantages in CCSP branchand- prune with local search of global optimization applied locally over each pruned branch of the CCSP. We apply local search techniques of continuous optimization over the pruned boxes outputted by the CCSP techniques. We mainly use steepest descent technique with different characteristics such as penalty calculation and step length. We implement two main different local search algorithms. We use “Procure”, which is a constraint reasoning and global optimization framework, to implement our techniques, then we produce and introduce our results over a set of benchmarks.

Probabilistic constraint reasoning with Monte Carlo integration to robot localization

This work studies the combination of safe and probabilistic reasoning through the hybridization of Monte Carlo integration techniques with continuous constraint programming. In continuous constraint programming there are variables ranging over continuous domains (represented as intervals) together with constraints over them (relations between variables) and the goal is to find values for those variables that satisfy all the constraints (consistent scenarios). Constraint programming “branch-and-prune” algorithms produce safe enclosures of all consistent scenarios. Special proposed algorithms for probabilistic constraint reasoning compute the probability of sets of consistent scenarios which imply the calculation of an integral over these sets (quadrature). In this work we propose to extend the “branch-and-prune” algorithms with Monte Carlo integration techniques to compute such probabilities. This approach can be useful in robotics for localization problems. Traditional approaches are based on probabilistic techniques that search the most likely scenario, which may not satisfy the model constraints. We show how to apply our approach in order to cope with this problem and provide functionality in real time.

The Triplaria tree (Triplaris spp) and Pseudomyrmex ants: a symbiotic relationship with risks of attack for humans

The authors report a massive attack by Pseudomyrmex ants on a human who touched a Triplaria - novice tree (Triplaris spp). The ants naturally live in these trees and their stings cause intense pain and discrete to moderate local inflammation. The problem is common in some Brazilian regions and can be prevented by identifying the trees.

Dynamic delta hedging of autocallables under a discrete rebalancing context

This work tests different delta hedging strategies for two products issued by Banco de Investimento Global in 2012. The work studies the behaviour of the delta and gamma of autocallables and their impact on the results when delta hedging with different rebalancing periods. Given its discontinuous payoff and path dependency, it is suggested the hedging portfolio is rebalanced on a daily basis to better follow market changes. Moreover, a mixed strategy is analysed where time to maturity is used as a criterion to change the rebalancing frequency.

On the pricing of bivariate options in the presence of a discrete dividend payment

A Work Project, presented as part of the requirements for the Award of a Master's Double Degree in Finance from the NOVA School of Business and Economics / Masters Degree in Economics from Insper

The accuracy of the escrowed dividend model on the value of European options on a stock paying discrete dividend

This project characterizes the accuracy of the escrowed dividend model on the value of European options on a stock paying discrete dividend. A description of the escrowed dividend model is provided, and a comparison between this model and the benchmark model is realized. It is concluded that options on stocks with either low volatility, low dividend yield, low ex-dividend to maturity ratio or that are deep in or out of the money are reasonably priced with the escrowed dividend model.

A public transport bus assignment problem: parallel metaheuristics assessment

Combinatorial Optimization Problems occur in a wide variety of contexts and generally are NP-hard problems. At a corporate level solving this problems is of great importance since they contribute to the optimization of operational costs. In this thesis we propose to solve the Public Transport Bus Assignment problem considering an heterogeneous fleet and line exchanges, a variant of the Multi-Depot Vehicle Scheduling Problem in which additional constraints are enforced to model a real life scenario. The number of constraints involved and the large number of variables makes impracticable solving to optimality using complete search techniques. Therefore, we explore metaheuristics, that sacrifice optimality to produce solutions in feasible time. More concretely, we focus on the development of algorithms based on a sophisticated metaheuristic, Ant-Colony Optimization (ACO), which is based on a stochastic learning mechanism. For complex problems with a considerable number of constraints, sophisticated metaheuristics may fail to produce quality solutions in a reasonable amount of time. Thus, we developed parallel shared-memory (SM) synchronous ACO algorithms, however, synchronism originates the straggler problem. Therefore, we proposed three SM asynchronous algorithms that break the original algorithm semantics and differ on the degree of concurrency allowed while manipulating the learned information. Our results show that our sequential ACO algorithms produced better solutions than a Restarts metaheuristic, the ACO algorithms were able to learn and better solutions were achieved by increasing the amount of cooperation (number of search agents). Regarding parallel algorithms, our asynchronous ACO algorithms outperformed synchronous ones in terms of speedup and solution quality, achieving speedups of 17.6x. The cooperation scheme imposed by asynchronism also achieved a better learning rate than the original one.

Snakebites as a largely neglected problem in the Brazilian Amazon: highlights of the epidemiological trends in the State of Amazonas

Envenoming snakebites are thought to be a particularly important threat to public health worldwide, especially in rural areas of tropical and subtropical countries. The true magnitude of the public health threat posed by snakebites is unknown, making it difficult for public health officials to optimize prevention and treatment. The objective of this work was to conduct a systematic review of the literature to gather data on snakebite epidemiology in the Amazon region and describe a case series of snakebites from epidemiological surveillance in the State of Amazonas (1974-2012). Only 11 articles regarding snakebites were found. In the State of Amazonas, information regarding incidents involving snakes is scarce. Historical trends show an increasing number of cases after the second half of the 1980s. Snakebites predominated among adults (20-39 years old; 38%), in the male gender (78.9%) and in those living in rural areas (85.6%). The predominant snake envenomation type was bothropic. The incidence reported by the epidemiological surveillance in the State of Amazonas, reaching up to 200 cases/100,000 inhabitants in some areas, is among the highest annual snakebite incidence rates of any region in the world. The majority of the cases were reported in the rainy season with a case-fatality rate of 0.6%. Snakebite envenomation is a great disease burden in the State of Amazonas, representing a challenge for future investigations, including approaches to estimating incidence under-notification and case-fatality rates as well as the factors related to severity and disabilities.

Geographic weighted regression: applicability to epidemiological studies of leprosy

Abstract: INTRODUCTION: Geographic information systems (GIS) enable public health data to be analyzed in terms of geographical variability and the relationship between risk factors and diseases. This study discusses the application of the geographic weighted regression (GWR) model to health data to improve the understanding of spatially varying social and clinical factors that potentially impact leprosy prevalence. METHODS: This ecological study used data from leprosy case records from 1998-2006, aggregated by neighborhood in the Duque de Caxias municipality in the State of Rio de Janeiro, Brazil. In the GWR model, the associations between the log of the leprosy detection rate and social and clinical factors were analyzed. RESULTS: Maps of the estimated coefficients by neighborhood confirmed the heterogeneous spatial relationships between the leprosy detection rates and the predictors. The proportion of households with piped water was associated with higher detection rates, mainly in the northeast of the municipality. Indeterminate forms were strongly associated with higher detections rates in the south, where access to health services was more established. CONCLUSIONS: GWR proved a useful tool for epidemiological analysis of leprosy in a local area, such as Duque de Caxias. Epidemiological analysis using the maps of the GWR model offered the advantage of visualizing the problem in sub-regions and identifying any spatial dependence in the local study area.

The effect of deforestation on the water cycle in the amazon basin: an attempt reformulate the problem

If widespread deforestation in Amazon results in reduced evaporative water flux, then either a decrease in evaporation is compensated locally by reduced rainfall,or else changed moisture balance expresses itself downwind in the yet undisturbed forest. The question of where rain will occur is crucial. It is suggested that the appearance of clouds and the occurrence of rainout is governed primarily by the interplay of local meteorologic and physical geography parameters with the atmospheric stability structure except for a few well-defined periods when rain is dominated by large scale atmospheric instability. This means that the study of these phenomena (local heat balances,studies on cloud formation mechanism, vertical atmospheric stability, etc.) must be made on the scale of the cloud size, a few tens of kilometers at most.

Uncertainty in olfactory decision-making

Relationships between accuracy and speed of decision-making, or speed-accuracy tradeoffs (SAT), have been extensively studied. However, the range of SAT observed varies widely across studies for reasons that are unclear. Several explanations have been proposed, including motivation or incentive for speed vs. accuracy, species and modality but none of these hypotheses has been directly tested. An alternative explanation is that the different degrees of SAT are related to the nature of the task being performed. Here, we addressed this problem by comparing SAT in two odor-guided decision tasks that were identical except for the nature of the task uncertainty: an odor mixture categorization task, where the distinguishing information is reduced by making the stimuli more similar to each other; and an odor identification task in which the information is reduced by lowering the intensity over a range of three log steps. (...)