arXiv subjects

Dao Nguyen

Publications and source records attributed to Dao Nguyen.

27 records · Page 2Linked to original sources

arXiv2020

Optimization and Discrete Approximation of Sweeping Processes with Controlled Moving Sets and Perturbations

This paper addresses a new class of optimal control problems for perturbed sweeping processes with measurable controls in additive perturbations of the dynamics and smooth controls in polyhedral moving sets. We develop a constructive discrete approximation procedure that allows us to strongly approximate any feasible trajectory of the controlled sweeping process by feasible discrete trajectories and also establish a $W^{1,2}$-strong convergence of optimal trajectories for discretized control problems to a given local minimizer of the original continuous-time sweeping control problem of the Bolza type. Employing advanced tools of first-order and second-order variational analysis and generalized differentiation, we derive necessary optimality conditions for discrete optimal solutions under fairly general assumptions formulated entirely in terms of the given data. The obtained results give us efficient suboptimality ("almost optimality") conditions for the original sweeping control problem that are illustrated by a nontrivial numerical example.

math.OC↗

arXiv2019

Automatic adaptation of MCMC algorithms

Markov chain Monte Carlo (MCMC) methods are ubiquitous tools for simulation-based inference in many fields but designing and identifying good MCMC samplers is still an open question. This paper introduces a novel MCMC algorithm, namely, Auto Adapt MCMC. For sampling variables or blocks of variables, we use two levels of adaptation where the inner adaptation optimizes the MCMC performance within each sampler, while the outer adaptation explores the valid space of kernels to find the optimal samplers. We provide a theoretical foundation for our approach. To show the generality and usefulness of the approach, we describe a framework using only standard MCMC samplers as candidate samplers and some adaptation schemes for both inner and outer iterations. In several benchmark problems, we show that our proposed approach substantially outperforms other approaches, including an automatic blocking algorithm, in terms of MCMC efficiency and computational time.

stat.ME↗

arXiv2019

A new Gini correlation between quantitative and qualitative variables

We propose a new Gini correlation to measure dependence between a categorical and numerical variables. Analogous to Pearson $R^2$ in ANOVA model, the Gini correlation is interpreted as the ratio of the between-group variation and the total variation, but it characterizes independence (zero Gini correlation mutually implies independence). Closely related to the distance correlation, the Gini correlation is of simple formulation by considering the nature of categorical variable. As a result, the proposed Gini correlation has a lower computational cost than the distance correlation and is more straightforward to perform inference. Simulation and real applications are conducted to demonstrate the advantages.

stat.ME↗

arXiv2019

Estimating Feature-Label Dependence Using Gini Distance Statistics

Identifying statistical dependence between the features and the label is a fundamental problem in supervised learning. This paper presents a framework for estimating dependence between numerical features and a categorical label using generalized Gini distance, an energy distance in reproducing kernel Hilbert spaces (RKHS). Two Gini distance based dependence measures are explored: Gini distance covariance and Gini distance correlation. Unlike Pearson covariance and correlation, which do not characterize independence, the above Gini distance based measures define dependence as well as independence of random variables. The test statistics are simple to calculate and do not require probability density estimation. Uniform convergence bounds and asymptotic bounds are derived for the test statistics. Comparisons with distance covariance statistics are provided. It is shown that Gini distance statistics converge faster than distance covariance statistics in the uniform convergence bounds, hence tighter upper bounds on both Type I and Type II errors. Moreover, the probability of Gini distance covariance statistic under-performing the distance covariance statistic in Type II error decreases to 0 exponentially with the increase of the sample size. Extensive experimental results are presented to demonstrate the performance of the proposed method.

cs.LG↗

arXiv2018

Simulation-based inference methods for partially observed Markov model via the R package is2

Partially observed Markov process (POMP) models are powerful tools for time series modeling and analysis. Inherited the flexible framework of R package pomp, the is2 package extends some useful Monte Carlo statistical methodologies to improve on convergence rates. A variety of efficient statistical methods for POMP models have been developed including fixed lag smoothing, second-order iterated smoothing, momentum iterated filtering, average iterated filtering, accelerate iterated filtering and particle iterated filtering. In this paper, we show the utility of these methodologies based on two toy problems. We also demonstrate the potential of some methods in a more complex model, employing a nonlinear epidemiological model with a discrete population, seasonality, and extra-demographic stochasticity. We discuss the extension beyond POMP models and the development of additional methods within the framework provided by is2.

stat.CO↗

arXiv2018

Optimal Control of Sweeping Processes in Robotics and Traffic Flow Models

The paper is mostly devoted to applications of a novel optimal control theory for perturbed sweeping/Moreau processes to two practical dynamical models. The first model addresses mobile robot dynamics with obstacles, and the second one concerns control and optimization of traffic flows. Describing these models as controlled sweeping processes with pointwise/hard control and state constraints and applying new necessary optimality conditions for such systems allow us to develop efficient procedures to solve naturally formulated optimal control problems for the models under consideration and completely calculate optimal solutions in particular situations.

math.OC↗

arXiv2018

Optimization of a perturbed sweeping process by constrained discontinuous controls

This paper deals with optimal control problems described by a controlled version of Moreau's sweeping process governed by convex polyhedra, where measurable control actions enter additive perturbations. This class of problems, which addresses unbounded discontinuous differential inclusions with intrinsic state constraints, is truly challenging and underinvestigated in control theory while being highly important for various applications. To attack such problems with constrained measurable controls, we develop a refined method of discrete approximations with establishing its well-posedness and strong convergence. This approach, married to advanced tools of first-order and second-order variational analysis and generalized differentiations, allows us to derive adequate collections of necessary optimality conditions for local minimizers, first in discrete-time problems and then in the original continuous-time controlled sweeping process by passing to the limit. The new results include an appropriate maximum condition and significantly extend the previous ones obtained under essentially more restrictive assumptions. We compare them with other versions of the maximum principle for controlled sweeping processes that have been recently established for global minimizers in problems with smooth sweeping sets by using different techniques. The obtained necessary optimality conditions are illustrated by several examples.

math.OC↗

arXiv2018

Accelerate iterated filtering

In simulation-based inferences for partially observed Markov process models (POMP), the by-product of the Monte Carlo filtering is an approximation of the log likelihood function. Recently, iterated filtering [14, 13] has originally been introduced and it has been shown that the gradient of the log likelihood can also be approximated. Consequently, different stochastic optimization algorithm can be applied to estimate the parameters of the underlying models. As accelerated gradient is an efficient approach in the optimization literature, we show that we can accelerate iterated filtering in the same manner and inherit that high convergence rate while relaxing the restricted conditions of unbiased gradient approximation. We show that this novel algorithm can be applied to both convex and non-convex log likelihood functions. In addition, this approach has substantially outperformed most of other previous approaches in a toy example and in a challenging scientific problem of modeling infectious diseases.

stat.ME↗

arXiv2015

Statistical Inference for Partially Observed Markov Processes via the R Package pomp

Partially observed Markov process (POMP) models, also known as hidden Markov models or state space models, are ubiquitous tools for time series analysis. The R package pomp provides a very flexible framework for Monte Carlo statistical investigations using nonlinear, non-Gaussian POMP models. A range of modern statistical methods for POMP models have been implemented in this framework including sequential Monte Carlo, iterated filtering, particle Markov chain Monte Carlo, approximate Bayesian computation, maximum synthetic likelihood estimation, nonlinear forecasting, and trajectory matching. In this paper, we demonstrate the application of these methodologies using some simple toy problems. We also illustrate the specification of more complex POMP models, using a nonlinear epidemiological model with a discrete population, seasonality, and extra-demographic stochasticity. We discuss the specification of user-defined models and the development of additional methods within the programming environment provided by pomp.

stat.ME↗