检索结果-内蒙古大学图书馆

Learning Bayesian network parameters via minimax algorithm

INTERNATIONAL JOURNAL OF APPROXIMATE REASONING 2019年 108卷 62-75页

作者： Gao, Xiao-guang Guo, Zhi-gao Ren, Hao Yang, Yu Chen, Da-qing He, Chu-chao Northwestern Polytech Univ Dept Syst Engn Xian Shaanxi Peoples R China London South Bank Univ Div Comp Sci & Informat London England

Parameter learning is an important aspect of learning in Bayesian networks. Although the maximum likelihood algorithm is often effective, it suffers from overfitting when there is insufficient data. To address this, prior distributions of model parameters are often imposed. When training a Bayesian network, the parameters of the network are optimized to fit the data. However, imposing prior distributions can reduce the fitness between parameters and data. Therefore, a trade-off is needed between fitting and overfitting. In this study, a new algorithm, named minimax Fitness (MMF) is developed to address this problem. The method includes three main steps. First, the maximum a posterior estimation that combines data and prior distribution is derived. Then, the hyper-parameters of the prior distribution are optimized to minimize the fitness between posterior estimation and data. Finally, the order of posterior estimation is checked and adjusted to match the order of the statistical counts from the data. In addition, we introduce an improved constrained maximum entropy method, named Prior Free Constrained Maximum Entropy (PF-CME), to facilitate parameter learning when domain knowledge is provided. Experiments show that the proposed methods outperforms most of existing parameter learning methods. (C) 2019 Elsevier Inc. All rights reserved.

关键词： Bayesian network Overfitting minimax algorithm Maximum entropy

来源：评论

学校读者我要写书评

暂无评论

An accelerated minimax algorithm for convex-concave saddle point problems with nonsmooth coupling function

引用

COMPUTATIONAL OPTIMIZATION AND APPLICATIONS 2023年第3期86卷 925-966页

作者： Bot, Radu Ioan Csetnek, Erno Robert Sedlmayer, Michael Univ Vienna Fac Math Vienna Austria Univ Vienna Res Network Data Sci Uni Vienna Vienna Austria

In this work we aim to solve a convex-concave saddle point problem, where the convex-concave coupling function is smooth in one variable and nonsmooth in the other and not assumed to be linear in either. The problem is augmented by a nonsmooth regulariser in the smooth component. We propose and investigate a novel algorithm under the name of OGAProx, consisting of an optimistic gradient ascent step in the smooth variable coupled with a proximal step of the regulariser, and which is alternated with a proximal step in the nonsmooth component of the coupling function. We consider the situations convex-concave, convex-strongly concave and strongly convex-strongly concave related to the saddle point problem under investigation. Regarding iterates we obtain (weak) convergence, a convergence rate of order O(1/K) and linear convergence like O(theta(K)) with theta < 1, respectively. In terms of function values we obtain ergodic convergence rates of order O(1/K), O(1/K-2) and O(theta(K)) with theta < 1, respectively. We validate our theoretical considerations on a nonsmooth-linear saddle point problem, the training of multi kernel support vector machines and a classification problem incorporating minimax group fairness.

关键词： Saddle point problem Convex-concave minimax algorithm Convergence rate Acceleration Linear convergence

来源：评论

学校读者我要写书评

暂无评论

Development of solar radiation prediction model using minimax algorithm

引用

ENVIRONMENTAL PROGRESS & SUSTAINABLE ENERGY 2024年第2期43卷 e14276-e14276页

作者： Kaplan, Yusuf Alper Osmaniye Korkut Ata Univ Dept Energy Syst Engn Osmaniye Turkiye Osmaniye Korkut Ata Univ Dept Energy Syst Engn TR-80000 Osmaniye Turkiye

Recently, many studies are carried out on solar energy (SE) and global solar radiation (GSR) estimation. When the literature is examined, there are many GSR estimation models developed with different algorithms in different regions of the world. Although some of these models are seen to perform acceptable all over the world, the performance of some models differs according to their location. The minimax algorithm is a decision-making algorithm that maximizes the winning potential while minimizing the probability of losing. In this study, a new Angstorm model was developed for GSR with minimax algorithm. The most important aspect of this article in terms of originality is that the GSR estimation model is being developed with the minimax algorithm for the first time. The MATLAB program was used while developing the prediction model with the minimax algorithm. Different statistical error tests were used to evaluate the performance of the results. When these test results are evaluated in general, it is seen that the minimax algorithm gives acceptable results in GSR estimation.

关键词： minimax algorithm model developing solar radiation statistical analysis

来源：评论

学校读者我要写书评

暂无评论

IMPROVED minimax algorithm FOR CIRCUIT OPTIMIZATION

引用

ELECTRONICS LETTERS 1980年第11期16卷 436-437页

作者： VIDALLON, CM Laboratoire de Genie Electrique Université Paul Sabatier Toulouse France

An algorithm is proposed to solve minimax problems, based on the processing of a locally linearised subproblem, with a strictly steepest descent procedure.

关键词： minimax techniques Optimisation techniques Computer-aided circuit analysis and design circuit CAD minimax algorithm circuit optimisation

来源：评论

学校读者我要写书评

暂无评论

Arbitrary antenna array pattern synthesis using minimax algorithm

引用

ELECTRONICS LETTERS 2001年第4期37卷 206-208页

作者： Dotlic, ID Zejak, AJ IMTEL YU-11070 Belgrade Yugoslavia

A novel approach to arbitrary array pattern synthesis is proposed. It is based on the minimax algorithm originally developed for mismatched filter design for radars with pulse compression. The proposed approach allows... 详细信息

关键词： minimax techniques Antenna theory Optimisation techniques antenna array pattern synthesis antenna radiation patterns antenna arrays minimax algorithm Antenna arrays

来源：评论

学校读者我要写书评

暂无评论

A Novel minimax algorithm for Multi-channel Active Noise Control System 12

A Novel Minimax Algorithm for Multi-channel Active Noise Con...

引用

12th International Symposium on Image and Signal Processing and Analysis (ISPA)

作者： Jain, Manish Kumar, Arun Bahl, Rajendar Indian Inst Technol Delhi Ctr Appl Res Elect New Delhi India

ISBN: (纸本)9781665426398

Global active noise control (ANC) employs multichannel filtered-x least mean square (MCFxLMS) algorithm as it is more suitable algorithm to obtain large quiet zone. minimax algorithm was proposed to counter the higher computational complexity faced in MCFxLMS based ANC by minimizing the square of the maximum of the absolute values of residual noise at the error microphones. However, the minimax approach leads to inferior performance in terms of convergence as well as noise reduction. Also, the classical minimax approach offers little flexibility in adjusting the ANC performance. In this paper, a novel minimax algorithm is proposed in order to tackle these shortcomings of conventional minimax algorithm at a cost of increase in computational complexity as compared to conventional minimax algorithm. The performance of the proposed approach is evaluated and compared with classical minimax for global noise reduction in a 2-dimensional quiet zone of size 1 m x 1 m in a 3-dimensional reverberant room. The proposed scheme is able to improve the performance with much reduced computational complexity as compared to MCFxLMS though with increased computational complexity as compared to classical minimax approach.

关键词： Active noise control multi-channel FxLMS minimax algorithm global noise reduction

来源：评论

学校读者我要写书评

暂无评论

Dismath: Discrete Math Gamified Educational Tool in Logic with minimax algorithm and Pruning 11

Dismath: Discrete Math Gamified Educational Tool in Logic wi...

引用

11th IEEE International Conference on Humanoid, Nanotechnology, Information Technology, Communication and Control, Environment, and Management ( HNICEM )

作者： Cabatuan, Melvin De La Salle Univ Elect & Commun Engn Dept Manila Philippines

ISBN: (纸本)9781728130446

This paper presents the design and development of Dismath, a gamified educational tool for propositional logic introduction to undergraduate computer engineering students. Additionally, an artificial intelligence (AI) agent is developed for the proposed Dismath game based on minimax tree search with alpha-beta pruning (MMAB). Dismath is an international checker variant inspired by Damath game in which operations are defined in every checkerboard positions. Thus, the final score is not only dependent on the pieces left but also in the operation results. In a capture move, the jumping chip is the first operand, the captured piece acts as the second operand, and the operation defined in the destination block where the chip landed after the jump is utilized for the binary operation. Instead of arithmetic operators, Dismath designates logical connectives on the checkerboard. All white and black pieces were assigned true and false truth values. For scoring, a true (T) and a false (F) truth values are calculated as +1 and -1, while a dama or king chip is valued +2 and -2 for White and Black pieces, respectively. If the game ends with a positive value, then white wins;otherwise black wins, unless the score is zero in which the outcome is draw. Board balancing experiments were conducted using the win-win ratio between the two random players as evaluation metric. Furthermore, the MMAB agent was characterized against a random player baseline. The results showed dominant performance of MMAB AI agent in proposed Dismath game losing only at depth equal to 1. Overall, this result is promising on its own demonstrating the automated gameplay of the proposed Dismath educational tool for logic introduction.

关键词： gamification propositional logic board games damath minimax algorithm alpha-beta pruning evolutionary

来源：评论

学校读者我要写书评

暂无评论

Improvisation of minimax algorithm with Multi Criteria Decision Maker (MCDM) in the Intelligent Agent of Card Battle Game 3

Improvisation of Minimax Algorithm with Multi Criteria Decis...

引用

3rd International Conference on Electrical Engineering and Computer Science (ICECOS)

作者： Permana, Silvester Dian Handy Bintoro, Ketut Bayu Yogha Arifitama, Budi Syahputra, Ade Cendana, Maya Univ Trilogi Informat Engn Study Program Jakarta Indonesia

ISBN: (纸本)9781728147147

Choosing the right algorithm for Intelligent Agent is very important in determining the next step. Actions taken by an Intelligent Agent with less optimum algorithm can be figured out easily by the player competing with it. This certainly makes it easier for the player to win the game. Card Battle is a game that compares the strength of both card decks. They competed to determine who has the best cards on deck to reduce the enemy health point. Intelligent Agent need to act as the competing enemy called Non-Player Character (NPC). Presently, Intelligent Agent on Card Battle Games still use the minimax algorithm. This algorithm thinks only to counter enemy summoned card without considering other things such as enemy health point left, the cost for summoning a card, etc. Improvisation of minimax algorithm with Multi Criteria Decision Maker (MCDM) provides a new way of thinking for the decision making during the game. The results of this study increase the difficulty of the NPC by making it possible for the Intelligent Agent to consider more game properties and counter the player cards.

关键词： Improvement algorithm minimax algorithm MCDM Intelligent Agent Card Battle Games

来源：评论

学校读者我要写书评

暂无评论

A PARTIAL-UPDATE minimax algorithm FOR PRACTICAL IMPLEMENTATION OF MULTI-CHANNEL FEEDFORWARD ACTIVE NOISE CONTROL 16

A PARTIAL-UPDATE MINIMAX ALGORITHM FOR PRACTICAL IMPLEMENTAT...

引用

16th International Workshop on Acoustic Signal Enhancement (IWAENC)

作者： Shi, Chuang Kajikawa, Yoshinobu Univ Elect Sci & Technol China Sch Informat & Commun Engn Chengdu Sichuan Peoples R China Kansai Univ Fac Engn Sci Osaka Japan

ISBN: (纸本)9781538681510

In the implementation of a multi-channel feedforward active noise control (ANC) system, the standard FxLMS algorithm demands too much computational power to be real-time processed by a low-cost digital processor. Therefore, various algorithms have been proposed to reduce their computational complexity by modifying the cost function and introducing the partial update (P-U). This paper proposes a new P-U minimax algorithm, which minimizes the uniform-norm instead of the 2-norm of the error signals and updates the filtered reference signals partially. The P-U minimax algorithm is compared with existing algorithms based on an experimental setup of a case (1, 4, 8) ANC system. The proposed P-U minimax algorithm is validated to be efficient and outperform the scanning error and mixed-error algorithms with the same degree of computational complexity.

关键词： Active noise control minimax algorithm partial-update adaptive filter

来源：评论

学校读者我要写书评

暂无评论

Phase Array Antenna Pattern Design Using the minimax algorithm

Phase Array Antenna Pattern Design Using the Minimax Algorit...

引用

2011 IEEE CIE International Conference on Radar(2011年IEEE国际雷达会议RADAR 2011)

作者： Meng Qian James Jen Zekeriya Aliyazicioglu H. K. Hwang Electrical and Computer Engineering California State Polytechnic University-Pomona 3801 W. Temple A Electrical and Computer Engineering California State Polytechnic University-Pomona 3801 W. Temple A

This paper investigates the performance of beam pattern forming employing the minimax algorithm. The examples used in this paper demonstrate how phase-only adjustment is performed on both circular and elliptical beams. Compared with complex weights adjustment, problem of phase only adjustment is addressed and performance analysis is provided. A 244 element non-uniformly spaced array antenna is used in this simulation study.

关键词： Array antenna minimax algorithm beamforming advanced signal processing.

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

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

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

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

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：