检索结果-内蒙古大学图书馆

Imitation Learning for Adaptive Video Streaming With Future Adversarial Information Bottleneck Principle

IEEE TRANSACTIONS ON MOBILE COMPUTING 2024年第12期23卷 13670-13683页

作者： Wang, Shuoyao Lin, Jiawei Ye, Fangwei Shenzhen Univ Coll Elect & Informat Engn Shenzhen 518060 Peoples R China Nanjing Univ Aeronaut & Astronaut Coll Comp Sci & Technol Nanjing 210095 Peoples R China

Adaptive video streaming plays a crucial role in ensuring high-quality video streaming services. Despite extensive research efforts devoted to Adaptive BitRate (ABR) techniques, the current reinforcement learning (RL)-based ABR algorithms may benefit the average Quality of Experience (QoE) but suffers from fluctuating performance in individual video sessions. In this paper, we present a novel approach that combines imitation learning with the information bottleneck technique, to learn from the complex offline optimal scenario rather than inefficient exploration. In particular, we leverage the deterministic offline bitrate optimization problem with the future throughput realization as the expert and formulate it as a mixed-integer non-linear programming (MINLP) problem. To enable large-scale training for improved performance, we propose an alternative optimization algorithm that efficiently solves the formulated MINLP problem. To address the overfitting issues due to the future information leakage in MINLP, we incorporate an adversarial information bottleneck framework. By compressing the video streaming state into a latent space, we retain only action-relevant information. Additionally, we introduce a future adversarial term to mitigate the influence of future information leakage, where Model Prediction Control (MPC) policy without any future information is employed as the adverse expert. Experimental results demonstrate the effectiveness of our proposed approach in significantly enhancing the quality of adaptive video streaming, providing a 7.30% average QoE improvement and a 30.01% average ranking reduction.

关键词： Adaptive video streaming imitation learning information bottleneck mixed-integer non-linear programming Adaptive video streaming imitation learning information bottleneck mixed-integer non-linear programming

来源：评论

学校读者我要写书评

暂无评论

On global optimization with indefinite quadratics

引用

EURO JOURNAL ON COMPUTATIONAL OPTIMIZATION 2017年第3期5卷 309-337页

作者： Fampa, Marcia Lee, Jon Melo, Wendel Univ Fed Rio de Janeiro Rio De Janeiro Brazil Univ Michigan Ann Arbor MI 48109 USA

We present an algorithmic framework for global optimization problems in which the non-convexity is manifested as an indefinite-quadratic as part of the objective function. Our solution approach consists of applying a spatial branch-and-bound algorithm, exploiting convexity as much as possible, not only convexity in the constraints, but also extracted from the indefinite-quadratic. A preprocessing stage is proposed to split the indefinite-quadratic into a difference of convex quadratic functions, leading to a more efficient spatial branch-and-bound focused on the isolated non-convexity. We investigate several natural possibilities for splitting an indefinite quadratic at the preprocessing stage, and prove the equivalence of some of them. Through computational experiments with our new solver iquad, we analyze how the splitting strategies affect the performance of our algorithm, and find guidelines for choosing amongst them.

关键词： Global optimization Indefinite quadratic Difference of convex functions Eigendecomposition Semidefinite programming mixed-integer non-linear programming

来源：评论

学校读者我要写书评

暂无评论

Complexity of transmission network expansion planning NP-hardness of connected networks and MINLP evaluation

引用

ENERGY SYSTEMS-OPTIMIZATION MODELING SIMULATION AND ECONOMIC ASPECTS 2014年第1期5卷 179-207页

作者： Oertel, David Ravi, R. Karlsruhe Inst Technol Karlsruhe Germany Carnegie Mellon Univ Tepper Sch Business Pittsburgh PA 15213 USA

Transmission network expansion planning in its original formulation is NP-hard due to the subproblem Steiner trees, the minimum cost connection of an initially unconnected network with mandatory and optional nodes. By using electrical network theory we show why NP-hardness still holds when this subproblem of network design from scratch is omitted by considering already (highly) connected networks only. This refers to the case of extending a long working transmission grid for increased future demand. It will be achieved by showing that this case is computationally equivalent to 3-SAT. Additionally, the original mathematical formulation is evaluated by using an appropriate state-of-the-art mixed integer non-linear programming solver in order to see how much effort in computation and implementation is really necessary to solve this problem in practice.

关键词： Transmission network expansion planning NP-hard mixed-integer non-linear programming Electrical network

来源：评论

学校读者我要写书评

暂无评论

A novel hybrid approach to allocate renewable energy sources in distribution system

引用

SUSTAINABLE ENERGY TECHNOLOGIES AND ASSESSMENTS 2015年 10卷 1-11页

作者： Singh, A. K. Parida, S. K. IITRAM Dept Elect Engn Ahmadabad Gujarat India Indian Inst Technol Dept Elect Engn Patna Bihar India

The integration of renewable energy sources (RESs) based DGs into grid has a great importance in improving system reliability. Many methods were proposed in the literature for finding best locations for DG placement considering various criteria. Sometime, it becomes difficult for combined placement of different kinds of renewable based DGs, such as solar, wind and fuel cell. The criterion of minimizing total system cost was used previously by many researchers for locating the optimal sites for DGs using OPF formulations. In this case, three different cost functions are formulated for different kinds of renewable energy sources (RESs). By taking combined cost function of all the RESs in the OPF to identify location for each different kind of sources becomes very cumbersome task. It would be difficult to find the exact locations for various kinds of RESs that is where to place which type of RESs. In order to solve this difficulty, three different objectives have been considered separately for determining the optimal locations for each kind of RESs using mixed integer nonlinear programming (MINLP) method considering loadability, losses and cost. Having many alternatives with these three objectives, analytic hierarchy process (AHP) has been used to make a decision over getting the optimal locations for these different kinds of RESs. The proposed methodology has been demonstrated on 15- node radial distribution system and 69-node mesh distribution system. (C) 2015 Elsevier Ltd. All rights reserved.

关键词： Analytical hierarchical process Distributed generation mixed-integer non-linear programming Optimal power flow Renewable energy sources

来源：评论

学校读者我要写书评

暂无评论

An MISOCP-Based Solution Approach to the Reactive Optimal Power Flow Problem

引用

IEEE TRANSACTIONS ON POWER SYSTEMS 2021年第1期36卷 529-532页

作者： Kayacik, Sezen Ece Kocuk, Burak Sabanci Univ Ind Engn Program TR-34956 Istanbul Turkey

In this letter, we present an alternative mixed-integer non-linear programming formulation of the reactive optimal power flow (ROPF) problem. We utilize a mixed-integer second-order cone programming (MISOCP) based approach to find global optimal solutions of the proposed ROPF problem formulation. We strengthen the MISOCP relaxation via the addition of convex envelopes and cutting planes. Computational experiments on challenging test cases show that the MISOCP-based approach yields promising results compared to a semidefinite programming based approach from the literature.

关键词： mixed-integer non-linear programming reactive optimal power flow second-order cone programming

来源：评论

学校读者我要写书评

暂无评论

Integrated planning for public health emergencies: A modified model for controlling H1N1 pandemic

引用

JOURNAL OF THE OPERATIONAL RESEARCH SOCIETY 2020年第5期71卷 748-761页

作者： Liu, Ming Xu, Xifen Cao, Jie Zhang, Ding Nanjing Univ Sci & Technol Sch Econ & Management 200 Xiaolingwei Rd Nanjing 210094 Jiangsu Peoples R China Nanjing Univ Informat Sci & Technol Reading Acad Nanjing Jiangsu Peoples R China SUNY Coll Oswego Sch Business Oswego NY 13126 USA

Infectious disease outbreaks have occurred many times in the past decades and are more likely to occur in the future. Recently, Buyuktahtakin et al. (2018) proposed a new epidemics-logistics model to control the 2014 Ebola outbreak in West Africa. Considering that different diseases have dissimilar diffusion dynamics and can cause different public health emergencies, we modify the proposed model by changing capacity constraint, and then apply it to control the 2009 H1N1 outbreak in China. We formulate the problem to be a mixed-integer non-linear programming model (MINLP) and simultaneously determine when to open the new isolated wards and when to close the unused isolated wards. The test results reveal that our model could provide effective suggestions for controlling the H1N1 outbreak, including the appropriate capacity setting and the minimum budget required with different intervention start times.

关键词： Epidemic outbreak resource allocation mixed-integer non-linear programming public health emergencies optimisation

来源：评论

学校读者我要写书评

暂无评论

Supporting supply chain planning and scheduling decisions in the oil and chemical industry

引用

COMPUTERS & CHEMICAL ENGINEERING 2004年第6-7期28卷 863-870页

作者： Lasschuit, W Thijssen, N Shell Global Solut Int NL-2501 CM The Hague Netherlands

In the downstream oil and chemical industry, planning and scheduling are resource-intensive, complex, rolling processes. Decisions are taken at different stages within the supply chain (supply, manufacturing and distribution) and at different levels in the management hierarchy (planning, scheduling and operations). They differ in business scope, time horizon & resolution, data certainty & accuracy, process detail and optimizing mechanism. Aligning each step of this complex process is critical to competitive advantage. Decision support tools must therefore be provided within a coherent framework, including mechanisms which allow consistent economic and operational steering, taking due account of available (real-time) information on actual operations and market economics. At the strategic and global planning level for a network of manufacturing plants, decisions have to be taken on feedstock procurement & distribution, utilization of production capacities, utilization of modes of transport and demand allocation. Not only existing capabilities have to be considered, but also new opportunities in all areas have to be evaluated. The resulting mathematical programming model is a mixed-integer non-linear programming (MINLP) model: integer aspects arise because of e.g. fixed costs/investment costs, tiered pricing and cargo costs. non-linear relations are mainly caused by multiplication of quantity and economic variables. In the presentation, the various strategic planning problem areas, the contents of the MINLP aspects and the implemented solution approach will be further elaborated. The use of such models during the aforementioned (strategic) decision-taking process yields substantial benefits not only in economic terms but also in an improved understanding of the interactions between the various components of the business. (C) 2003 Published by Elsevier Ltd.

关键词： strategic business planning mixed-integer non-linear programming

来源：评论

学校读者我要写书评

暂无评论

Solving Pseudo-Convex mixed integer Optimization Problems by Cutting Plane Techniques

引用

OPTIMIZATION AND ENGINEERING 2002年第3期3卷 253-280页

作者： Westerlund, Tapio Porn, Ray Abo Akad Univ Proc Design Lab FIN-20500 Turku Finland Swedish Polytech Technol & Commun FIN-65200 Vaasa Finland

In the present paper a cutting plane approach to solve mixed-integer non-linear programming (MINLP) problems, containing pseudo-convex functions, is given. It is shown how valid cutting planes for pseudo convex functions can be obtained and, furthermore, it is shown how a class of non-convex MINLP problems with a pseudo-convex objective function and pseudo-convex constraints, can be solved to global optimality with the considered cutting plane technique. Finally the numerical efficiency of the procedure, when solving some example problems, is illustrated.

关键词： optimization mixed-integer non-linear programming integer non-linear programming quasi-convex functions pseudo-convex functions non-convex MINLP problems extended cutting plane method

来源：评论

学校读者我要写书评

暂无评论

A flow-to-equity approach to coordinate supply chain network planning and financial planning with annual cash outflows to an institutional investor

引用

Business Research 2016年第2期9卷 297-333页

作者： Steinrücke, Martin Albrecht, Wolfgang Faculty of Law and Economics University of Greifswald Friedrich-Loeffler-Str. 70 Greifswald 17489 Germany

A common side effect of cross-linked global economies is that well-positioned middle class companies are acquired by institutional investors, which formulate unreasonable return expectations in many cases. As a consequence, the resulting payouts are often not in line with business operations so that even world market leaders get into trouble or close down. In this context, we consider the case of a sanitary company, which had to manage the described situation after a business takeover. In order to coordinate the annual cash outflows to the investor with intra-organizational supply chain planning and financial planning, we propose a mixed-integer non-linear programming model that is based on the flow-to-equity discounted cash flow method. The objective is to maximize the present value of equity while determining annual cash outflows to the institutional investor during his engagement. As the decisions of the investor during his engagement influence possible operations of the company after his engagement, the residual value of equity (that influences the selling price) is taken into account. The modeling is based on cash flow series, which result from supply chain operations and restructuring on the one hand, and from financial transactions on the other. Financing is characterized by interest rates depending on the time period the credit starts, the credit period, the debt limit of the company and the current total debt. As the latter is a result of the optimization, non-linearity arises. Nevertheless, both the expected demand scenario and further randomly generated demand scenarios of the sanitary company could be solved to the optimum with the commercial optimization package GAMS 23.8/SCIP 2.1.1 within acceptable computation times, if capacity profiles are assigned to the locations to depict feasible and/or preferred capacity developments. © 2016, The Author(s).

关键词： Annual cash outflows to the investor Capacity profiles Company takeover Flow-to-equity method mixed-integer non-linear programming Supply chain design

来源：评论

学校读者我要写书评

暂无评论

Bi-Criteria Optimisation for Energy Harvesting Relaying: Formulation and Algorithm Implementation

Bi-Criteria Optimisation for Energy Harvesting Relaying: For...

引用

IEEE Global Communications Conference (GLOBECOM)

作者： AlQahtani, Dokhyl Chen, Yunfei Univ Warwick Sch Engn Coventry W Midlands England

ISBN: (纸本)9781665435406

This paper proposes a bi-criteria optimisation framework that maximises both the network rate and the harvested energy, which are contradictory objectives. Using the practical non-linear energy harvesting (EH) model, we jointly optimise relay selection (RS), power splitting (PS) and power allocation (PA). We decouple the relay selection variables from the other resource allocation variables to convert the original mixed-integer non-linear programming (MINLP) problem into a tractable problem. For PS and PA, the well-known epsilon-constraint method is applied to convert the bi-criteria problem into a convex problem. For RS, we propose a sub-optimal algorithm based on a selection order function with linear complexity. The simulation results indicate that the proposed schemes perform better than the benchmarks, drastically reducing computational complexity from exponential to polynomial.

关键词： Bi-Criteria optimisation epsilon-constraint method mixed-integer non-linear programming relay selection non-linear energy harvesting model

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：