Search

Model-based offline reinforcement learning methods (RL) have achieved state-of-the-art performance in many decision-making problems thanks to their sample efficiency and generalizability. Despite these advancements, existing model-based offline RL approaches either focus on theoretical studies without developing practical algorithms or rely on a restricted parametric policy space, thus not fully leveraging the advantages of an unrestricted policy space inherent to model-based methods. To address this limitation, we develop MoMA, a model-based mirror ascent algorithm with general function approximations under partial coverage of offline data. MoMA distinguishes itself from existing literature by employing an unrestricted policy class. In each iteration, MoMA conservatively estimates the value function by a minimization procedure within a confidence set of transition models in the policy evaluation step, then updates the policy with general function approximations instead of commonly-used parametric policy classes in the policy improvement step. Under some mild assumptions, we establish theoretical guarantees of MoMA by proving an upper bound on the suboptimality of the returned policy. We also provide a practically implementable, approximate version of the algorithm. The effectiveness of MoMA is demonstrated via numerical studies.

In applications such as biomedical studies, epidemiology, and social sciences, recurrent events often co-occur with longitudinal measurements and a terminal event, such as death. Therefore, jointly modeling longitudinal measurements, recurrent events, and survival data while accounting for their dependencies is critical. While joint models for the three components exist in statistical literature, many of these approaches are limited by heavy parametric assumptions and scalability issues. Recently, incorporating deep learning techniques into joint modeling has shown promising results. However, current methods only address joint modeling of longitudinal measurements at regularly-spaced observation times and survival events, neglecting recurrent events. In this paper, we develop TransformerLSR, a flexible transformer-based deep modeling and inference framework to jointly model all three components simultaneously. TransformerLSR integrates deep temporal point processes into the joint modeling framework, treating recurrent and terminal events as two competing processes dependent on past longitudinal measurements and recurrent event times. Additionally, TransformerLSR introduces a novel trajectory representation and model architecture to potentially incorporate a priori knowledge of known latent structures among concurrent longitudinal variables. We demonstrate the effectiveness and necessity of TransformerLSR through simulation studies and analyzing a real-world medical dataset on patients after kidney transplantation.

Interface problems depict many fundamental physical phenomena and widely apply in the engineering. However, it is challenging to develop efficient fully decoupled numerical methods for solving degenerate interface problems in which the coefficient of a PDE is discontinuous and greater than or equal to zero on the interface. The main motivation in this paper is to construct fully decoupled numerical methods for solving nonlinear degenerate interface problems with ``double singularities". An efficient fully decoupled numerical method is proposed for nonlinear degenerate interface problems. The scheme combines deep neural network on the singular subdomain with finite difference method on the regular subdomain. The key of the new approach is to split nonlinear degenerate partial differential equation with interface into two independent boundary value problems based on deep learning. The outstanding advantages of the proposed schemes are that not only the convergence order of the degenerate interface problems on whole domain is determined by the finite difference scheme on the regular subdomain, but also can calculate $\mathbf{VERY}$ $\mathbf{BIG}$ jump ratio(such as $10^{12}:1$ or $1:10^{12}$) for the interface problems including degenerate and non-degenerate cases. The expansion of the solutions does not contains any undetermined parameters in the numerical method. In this way, two independent nonlinear systems are constructed in other subdomains and can be computed in parallel. The flexibility, accuracy and efficiency of the methods are validated from various experiments in both 1D and 2D. Specially, not only our method is suitable for solving degenerate interface case, but also for non-degenerate interface case. Some application examples with complicated multi-connected and sharp edge interface examples including degenerate and nondegenerate cases are also presented.

We study an optimal control problem governed by elliptic PDEs with interface, which the control acts on the interface. Due to the jump of the coefficient across the interface and the control acting on the interface, the regularity of solution of the control problem is limited on the whole domain, but smoother on subdomains. The control function with pointwise inequality constraints is served as the flux jump condition which we called Neumann interface control. We use a simple uniform mesh that is independent of the interface. The standard linear finite element method can not achieve optimal convergence when the uniform mesh is used. Therefore the state and adjoint state equations are discretized by piecewise linear immersed finite element method (IFEM). While the accuracy of the piecewise constant approximation of the optimal control on the interface is improved by a postprocessing step which possesses superconvergence properties; as well as the variational discretization concept for the optimal control is used to improve the error estimates. Optimal error estimates for the control, suboptimal error estimates for state and adjoint state are derived. Numerical examples with and without constraints are provided to illustrate the effectiveness of the proposed scheme and correctness of the theoretical analysis., Comment: 31pages, 12 figures, 4 tables

To respond to the digital building environment and Industry 4.0 in building information modeling (BIM)-based mechanical, electrical, and plumbing (MEP) systems this study developed a new approach to engineering capstone projects. This article reports on a case study of the development of the innovative capstone project for engineering majors, which is based on team-based learning (TBL) combined with the 360-degree evaluation feedback method to cultivate students' BIM competency. The data collection and analysis involves a combination of qualitative and quantitative methods, which aims to evaluate students' learning outcomes and their BIM competency. The results indicate that TBL, combined with the 360-degree evaluation feedback in the capstone project, can be highly effective in improving graduates' BIM competency. This study discusses the interoperability of BIM software tools, problems in the process of data exchange, and provides suggestions for improving the course and BIM team collaboration. The research has identified how professional capabilities for students can be enhanced, enabled through a capstone project, and educators are able to deploy the BIM course to develop engineers that closely meet industry needs. The study provides the case for using the capstone project to improve and cultivate engineering students' BIM competency in MEP systems. This study has provided a new paradigm for applying TBL and 360-degree evaluation feedback to engineering education.

The rapid growth of the Internet use has allowed social networks to become the most effective means for marketing, leading to the emergence of "viral marketing" as a business model. The biggest challenge that is facing "viral marketing" is selecting seed users from the whole user set to form a "seed-set" to spread the influence and maximize the number of influenced users. This is known as the classic influence maximization problem. Based on the background provided, the paper focus on the fairness of information Maximization in social media and try to explain how the homophily effect, rich get richer mechanism and the duration of top users impact users in temporal social network. The paper also aims at developing a time-Awareness disparity seeding framework based on Disparity seeding framework, which is proved by experiments to slove the absolute error problem that exist between the target ratio and influential ratio. Furthermore, the unequal seeding disperse algorithm (USD), equal seeding disperse algorithm(ESD) and origin seeding disperse algorithm(OSD) have been developed to improve the influence maximization in temporal social network. The purpose is to find the most cost-effective user seed-set in any given period of time to maximize the influence of target users and increase the number of influenced users. According to the experimental test result, it is found that unequal seeding disperse algorithm perform better than the other two algorithms. In this paper, many experiments are carried out to verify the effectiveness of all the three algorithms using real social network data sets. As a result, the effectiveness and efficiency of the proposed algorithm was proven., Computer Science

The rapid growth of the Internet use has allowed social networks to become the most effective means for marketing, leading to the emergence of "viral marketing" as a business model. The biggest challenge that is facing "viral marketing" is selecting seed users from the whole user set to form a "seed-set" to spread the influence and maximize the number of influenced users. This is known as the classic influence maximization problem. Based on the background provided, the paper focus on the fairness of information Maximization in social media and try to explain how the homophily effect, rich get richer mechanism and the duration of top users impact users in temporal social network. The paper also aims at developing a time-Awareness disparity seeding framework based on Disparity seeding framework, which is proved by experiments to slove the absolute error problem that exist between the target ratio and influential ratio. Furthermore, the unequal seeding disperse algorithm (USD), equal seeding disperse algorithm(ESD) and origin seeding disperse algorithm(OSD) have been developed to improve the influence maximization in temporal social network. The purpose is to find the most cost-effective user seed-set in any given period of time to maximize the influence of target users and increase the number of influenced users. According to the experimental test result, it is found that unequal seeding disperse algorithm perform better than the other two algorithms. In this paper, many experiments are carried out to verify the effectiveness of all the three algorithms using real social network data sets. As a result, the effectiveness and efficiency of the proposed algorithm was proven., Computer Science

Environmental performance allows measuring the existing gap between countries regarding their environmental policy objectives. Improving environmental performance also allows countries to achieve their sustainable development goals. However, a systematic identification of factors influencing environmental performance constitutes a premise to improve it and such a review of factors has not been conducted in previous research. This paper develops a quantitative literature review of the factors influencing environmental performance in which a total of 84 journal papers were identified by keyword retrieval between 2004 and 2019. Literature metrological and literature content analyses are performed and two major research outcomes are obtained: first, a list of environmental performance influencing factors and a classification of the five main research streams related to environmental performance: enterprise, government, economy, technology and society. Second, building on the previous classification, a research agenda is proposed which points out current shortcomings and potential research directions for environmental performance research. The results of this piece of research provides a theoretical reference for improving environmental performance. It can also help countries target better environmental management practices when seeking global sustainable development. ·Highlights ·84 journal papers are analysed with metrological and content analysis. ·Grounded theory produces a list of factors affecting environmental performance. ·Environmental performance factors are classified in five relevant categories. ·These categories are: enterprise, government, economy, society and technology. ·A research agenda is proposed for meeting the 2030 Sustainable development goals.

Clustering, a fundamental activity in unsupervised learning, is notoriously difficult when the feature space is high-dimensional. Fortunately, in many realistic scenarios, only a handful of features are relevant in distinguishing clusters. This has motivated the development of sparse clustering techniques that typically rely on k-means within outer algorithms of high computational complexity. Current techniques also require careful tuning of shrinkage parameters, further limiting their scalability. In this paper, we propose a novel framework for sparse k-means clustering that is intuitive, simple to implement, and competitive with state-of-the-art algorithms. We show that our algorithm enjoys consistency and convergence guarantees. Our core method readily generalizes to several task-specific algorithms such as clustering on subsets of attributes and in partially observed data settings. We showcase these contributions thoroughly via simulated experiments and real data benchmarks, including a case study on protein expression in trisomic mice.