Heuristic search and pruning in polynomial constraints satisfaction

In this paper, we report how certain AI techniques can be used to speed up an algebraic algorithm for deciding the satisfiability of a system of polynomial equations, dis-equations, and inequalities. We begin by viewing the algebraic algorithm (Cylindrical Algebraic Decomposition) as a search procedure which searches for a solution of the given system. When viewed in this way, the algebraic algorithm is non-deterministic, in the sense that it can often achieve the same goal, but following different search paths requiring different amounts of computing times. Obviously one wishes to follow the least time-consuming path. However, in practice it is not possible to determine such an optimal path. Thus it naturally renders itself to the heuristic search techniques of AI. In particular we experimented with Best-First strategy. We also study a way to prune the search space, which proves to be useful especially when the given polynomial system is not satisfiable. The experimental results indicate that such AI techniques can often help in speeding up the algebraic method, sometimes dramatically. This revised version was published online in June 2006 with corrections to the Cover Date.     

Exploiting the deep structure of constraint problems

We introduce a technique for analyzing the behavior of sophisticated AI search programs working on realistic, large-scale problems. This approach allows us to predict where, in a space of problem instances, the hardest problems are to be found and where the fluctuations in difficulty are greatest. Our key insight is to shift emphasis from modelling sophisticated algorithms directly to modelling a search space that captures their principal effects. We compare our model's predictions with actual data on real problems obtained independently and show that the agreement is quite good. By systematically relaxing our underlying modelling assumptions we identify their relative contribution to the remaining error and then remedy it. We also discuss further applications of our model and suggest how this type of analysis can be generalized to other kinds of AI problems.     

基于资源约束和局部启发搜索的规划系统

针对如何充分发挥车间设备的资源潜力,完成企业下达的生产任务,使完成任务所需时间和生产成本尽可能的少和智能规划中资源规划所要解决的典型问题,提出了基于资源约束和局部搜索相结合的规划方法。并且在资源的分配上采用了优先级分派原则,在局部搜索中通过约束表达引导搜索的启发式信息,施行规则控制下的启发式搜索,提高了车间资源规划的效率。根据上述恩想,结合顺特电气线圈生产车间资源规划的实际问题。研究开发了车间资源规划系统,实际运行例子表明,所提出的思想是有实用价值的。     

Agent在智能信息检索中的应用研究

从信息检索的现状出发 ,运用人工智能的技术 ,提出了一种基于Agent的信息检索方法。通过建立用户个性化信息表 ,辅助过滤Agent在信息提取过程中的精确性 ,从而弥补了现有系统在自适应用户兴趣和交互方式等方面的不足     

Vision-based robotics: a challenge to real world Artificial Intelligence

Vision is a key function not only for robotics but also for AI more generally. Today real-time visual processing is becoming possible; this means that vision based behavior can become more dynamic, opening fertile areas for applications. One aspect of this is real-time visual tracking. We have built a real-time tracking system and incorporated it in an integrated robot programming environment. Using this, we have performed experiments in vision based robot behavior and human-robot interactions. In particular, we have developed a robotic system capable of 'learning by seeing'. In general, it is important for the AI community not to lose sight of the problems and progress of robotics. After all, an AI system which acts in real-time in the real-world is no less (and no more) than an intelligent robot.     

Statistical Regimes Across Constrainedness Regions

Much progress has been made in terms of boosting the effectiveness of backtrack style search methods. In addition, during the last decade, a much better understanding of problem hardness, typical case complexity, and backtrack search behavior has been obtained. One example of a recent insight into backtrack search concerns so-called heavy-tailed behavior in randomized versions of backtrack search. Such heavy-tails explain the large variance in runtime often observed in practice. However, heavy-tailed behavior does certainly not occur on all instances. This has led to a need for a more precise characterization of when heavy-tailedness does and when it does not occur in backtrack search. In this paper, we provide such a characterization. We identify different statistical regimes of the tail of the runtime distributions of randomized backtrack search methods and show how they are correlated with the “sophistication” of the search procedure combined with the inherent hardness of the instances. We also show that the runtime distribution regime is highly correlated with the distribution of the depth of inconsistent subtrees discovered during the search. In particular, we show that an exponential distribution of the depth of inconsistent subtrees combined with a search space that grows exponentially with the depth of the inconsistent subtrees implies heavy-tailed behavior. Research supported by the Intelligent Information Systems Institute, Cornell University (AFOSR grant F49620-01-1-0076), MURI (AFOSR grant F49620-01-1-0361) and Ministerio de Educación y Ciencia (TIN-2004 grant 07933-C03-03). We thank the anonymous reviewers for their insightful comments.     

面向决策可托管应用的行为提取与分析方法

游戏平台为优化游戏体验常提供决策AI托管功能,这导致平台在对用户进行画像时面临数据质量低的问题。日志中混合了托管AI和玩家自主作出的决策行为,有时还存在行为缺失的情况。提出一种面向决策可托管应用的行为提取与分析方法,通过基于启发式规则的行为特征提取方法从日志中提取玩家自主行为。最后使用多种画像方法在原始数据集和精简数据集上分别进行分析并比较其结果,得出该方法可在大幅度缩减数据量的同时较好地保留玩家的行为特征信息。     

多媒体智能：当多媒体遇到人工智能

过去10年中涌现出大量新兴的多媒体应用和服务,带来了很多可以用于多媒体前沿研究的多媒体数据。多媒体研究在图像/视频内容分析、多媒体搜索和推荐、流媒体服务和多媒体内容分发等方向均取得了重要进展。与此同时,由于在深度学习领域所取得的重大突破,人工智能(artificial intelligence,AI)在20世纪50年代被正式视为一门学科之后,迎来了一次“新”的发展浪潮。因此,一个问题就自然而然地出现了：当多媒体遇到人工智能时会带来什么？为了回答这个问题,本文通过研究多媒体和人工智能之间的相互影响引入了多媒体智能的概念。从两个方面探讨多媒体与人工智能之间的相互影响：一是多媒体促使人工智能向着更具可解释性的方向发展;二是人工智能反过来为多媒体研究注入了新的思维方式。这两个方面形成了一个良性循环,多媒体和人工智能在其中不断促进彼此发展。本文对相关研究及进展进行了讨论,并围绕值得进一步探索的研究方向分享见解。希望可以对多媒体智能的未来发展带来新的研究思路。     

The use of a heuristic search technique for the optimization of quality of steel billets produced by continuous casting

This work proposes an interaction between a numerical heat transfer model and an artificial intelligence (AI) heuristic search method which is linked to a Knowledge Base for the continuous casting of quality billets (SAE 1010 steel). A two-dimensional heat transfer model was developed using a finite difference method and was applied to real dimensions of a steel billet continuous caster. The Knowledge Base consists of technological and metallurgical constraints. The heuristic search method, aided by the Knowledge Base, explores the space parameter settings in order to find optimized cooling conditions which results in defect free billet production with minimum metallurgical length. The use of the developed search method for determining optimum setting of cooling conditions demonstrated that very good results can be attained in assuring billet quality and in the improvement of caster performance.     

Using artificial intelligence for modeling of the realistic animal behaviors in a virtual island

The remarkable development of the computer graphic techniques enables the creation and management of more realistic games and virtual environments. However, placing the Artificial Intelligence (AI) in a virtual world get making these environments both more interactive and more believable. One of the most ambitious goals of the AI is to create virtual worlds in which a big number of virtual characters or humans being interacted and behave in an autonomous way. For this purpose, placing embodied intelligence characters in a virtual world offers a unique opportunity to evaluate the AI concept.This paper introduces a Virtual Island developed in an innovative way based on fuzzy rules at the user interaction mechanism. We have used fuzzy tactics to create AI-based animals having various behavior types from an eagles’ perspective which called ‘Flight Simulation’. The simulation utilizes an eagle flying over the airspace of the Island of Chios. The AI on the ground is triggered by other animals when they enter a radius area with a certain speed defined in the software. It then decides how to behave according to health, behavior type and confidence level. Also, there is non-AI sparrow herd placed over the island to make user understand how fast he is and give the user sense of speed what can be called as an in-project feature. Consequently, the used Fuzzy Tactics have been tested in realized Unity 3D simulation. The results of the study have proven that the virtual environment consisting AI-based animals has a good performance in terms of animal-user interactions and provided satisfactory results in run time.     

Putting AI in entertainment: an AI authoring tool for simulation and games

We have been working on techniques to write better AI behavior for simulations and games. We are focusing on a visual authoring tool that provides a way to quickly synthesize complex behavior, and are building a corresponding AI engine to run with a simulation or game. For game development uses, we see our tool as making the AI understandable to game designers and end users, as well as improving developers' productivity. This work will also be useful for simulation developers and subsequently for analysts, operators, and instructors. We explain some of our tool's features and how the AI engine processes the resulting content.     

Fast forward planning by guided enforced hill climbing

In recent years, a number of new heuristic search methods have been developed in the field of automated planning. Enforced hill climbing (EHC) is one such method which has been frequently used in a number of AI planning systems. Despite certain weaknesses, such as getting trapped in dead-ends in some domains, this method is more competitive than several other methods in many planning domains. In order to enhance the efficiency of ordinary enforced hill climbing, a new form of enforced hill climbing, called guided enforced hill climbing, is introduced in this paper. An adaptive branch ordering function is the main feature that guided enforced hill climbing has added to EHC. Guided enforced hill climbing expands successor states in the order recommended by the ordering function. Our experimental results in several planning domains show a significant improvement in the efficiency of the enforced hill climbing method, especially when applied to larger problems.     

Gout-ITS系统的深度优先语义遍历算法

该文分析了痛风临床诊治智能教学系统（Intelligent Tutoring System for the Instruction of Gout Clinical Diagnosis and Treatment，以下简称Gout-ITS系统）自动生成病例所需的领域知识及其特点，提出了语义树知识表示法和深度优先语义遍历算法。该算法可以有效地生成既符合学生的学习难度要求、又符合病理逻辑的、多样化不重复的病例。最后，将该算法与人工智能中的深度优先搜索算法进行了比较，阐述了其中的不同之处。     

面向Agent计算(AOC)原理

Ａｇｅｎｔ的出现使我们看到了ＡＩ进一步发展的希望,由于对Ａｇｅｎｔ缺乏统一的认识,使得在设计Ａｇｅｎｔ系统时出现一些混乱,我们认为造成这一状况的根本原因在于设计者对面向Ａｇｅｎｔ计算原理认识有偏差,本文将智能划分三个层次,并从知识层次和行为层次阐述ＡＯＣ基本原理。     

A case-based approach to heuristic planning

Most of the great success of heuristic search as an approach to AI Planning is due to the right design of domain-independent heuristics. Although many heuristic planners perform reasonably well, the computational cost of computing the heuristic function in every search node is very high, causing the planner to scale poorly when increasing the size of the planning tasks. For tackling this problem, planners can incorporate additional domain-dependent heuristics in order to improve their performance. Learning-based planners try to automatically acquire these domain-dependent heuristics using previous solved problems. In this work, we present a case-based reasoning approach that learns abstracted state transitions that serve as domain control knowledge for improving the planning process. The recommendations from the retrieved cases are used as guidance for pruning or ordering nodes in different heuristic search algorithms applied to planning tasks. We show that the CBR guidance is appropriate for a considerable number of planning benchmarks.     

ESTIMATION OF SOLID WASTE BIOMETHANIZATION: DEVELOPMENT OF NOVEL METHODOLOGY AND ANFIS MODELS

This article presents the application of a technique of artificial intelligence (AI) that explores the possibility of using a model to estimate the biomethanization of municipal solid waste (MSW). The model uses data from an experiment in which MSW is anaerobically digested under three different moisture regimes by leachate recycling. A method utilizing a neurofuzzy inference system is used because AI systems have a high capacity for empiric learning.

Considering the importance of finding an effective selection of the most valuable variables for the model, this methodology includes the following techniques: Exhaustive Search (or brute-force search); Stepwise, a step-by-step regression method; and the use of Expert Knowledge. With the use of the fuzzy logic toolbox (MATLAB®), nine models were generated. However, when a case study is used to detail the method, the proposed methodology can also be used with any other system with a set of input and output data.     

The challenges of real-time AI

The research agendas of artificial intelligence and real-time systems are converging as AI methods move toward domains that require real-time responses, and real-time systems move toward complex applications that require intelligent behavior. They meet at the crossroads in an exciting new subfield commonly called “real-time AI.” This subfield is still being defined, and the precise goals for various real-time AI systems are in flux. Our goal is to identify promising areas for future research in both real-time and AI techniques. We describe an organizing conceptual structure for current real-time AI research, exploring the meanings this term has acquired. We then identify the goals of real-time AI research and specify some necessary steps for reaching them     

An exact line search method for solving generalized continuous-timealgebraic Riccati equations

We present a Newton-like method for solving algebraic Riccati equations that uses an exact line search to improve the sometimes erratic convergence behavior of Newton's method. It avoids the problem of a disastrously large first step and accelerates convergence when Newton steps are too small or too long. The additional work to perform the line search is small relative to the work needed to calculate the Newton step     

A knowledge-based problem solving method in GIS application

Model design for theme analysis is one of the biggest challenges in GIS. Many real applications in GIS require functioning not only in data management and visualization, but also in analysis and decision-making. Confronted with an application of planning a new metro line in a city, a typical GIS is unable to accomplish the task in the absence of human experts or artificial intelligence technologies. Apart from being models for analyzing in different themes, some applications are also instances of problem solving in AI. Therefore, in order to strengthen its ability in automatic analysis, many theories and technologies from AI can be embedded in the GIS. In this paper, a state space is defined to formalize the metro line planning problem. By utilizing the defined state evaluation function, knowledge-based rules and strategies, a heuristic searching method is developed to optimize the solutions iteratively. Experiments are implemented to illuminate the validity of this AI-enhanced automatic analysis model of GIS.     

Survey of Mahjong game AI construction methods北大核心CSCD

Mahjong and its different variants have complex rules. Therefore, building a high-level Mahjong game artificial intelligence (AI) algorithm and its test environment is challenging. Through the analysis of relevant research literature on Mahjong game, this paper identified two types of Mahjong AI construction methods based on knowledge and data. Moreover, the advantages and disadvantages of each typical method are analyzed, emphasizing the construction method of Suphx. The problems and challenges encountered in constructing Mahjong AI are identified, suggesting the need to apply experience replay, hierarchical reinforcement learning, curiosity model, opponent model, metalearning, transfer learning, and curriculum learning to the AI algorithm optimization of Mahjong game and construct diversified Mahjong AI evaluation indicators, general confrontation platforms, and high-quality data sets. These problems are all promising research directions for the future. © 2023, Editorial Department of CAAI Transactions on Intelligent Systems. All rights reserved.