Walter D. Potter

An evaluation of local improvement operators for genetic algorithms

Genetic algorithms have demonstrated considerable success in providing good solutions to many NP-hard optimization problems. For such problems, exact algorithms that always find an optimal solution are only useful for small toy problems, so heuristic algorithms such as the genetic algorithm must be used in practice. In this paper, we apply the genetic algorithm to the NP-hard problem of multiple fault diagnosis (MFD). We compare a pure genetic algorithm with several variants that include local improvement operators. These operators, which are often domain-specific, are used to accelerate the genetic algorithm in converging on optimal solutions. Our empirical results indicate that by using the appropriate local improvement operator, the genetic algorithm is able to find an optimal solution in all but a tiny fraction of the cases and at a speed orders of magnitude faster than exact algorithms. >

An edge detection technique using genetic algorithm-based optimization

Abstract In this paper we present a genetic algorithm-based optimization technique for edge detection. The problem of edge detection is formulated as one of choosing a minimum cost edge configuration. The edge configurations are viewed as two-dimensional chromosomes with fitness values inversely proportional to their costs. The design of the crossover and the mutation operators in the context of the two-dimensional chromosomal representation is described. The knowledge-augmented mutation operator which exploits knowledge of the local edge structure is shown to result in rapid convergence. The incorporation of meta-level operators and strategies such as the elitism strategy, the engineered conditioning operator and adaptation of mutation and crossover rates in the context of edge detection are discussed and are shown to improve the convergence rate. The genetic algorithm with various combinations of meta-level operators is tested on synthetic and natural images. The performance of the genetic algorithm-based cost minimization technique is compared both qualitatively and quantitatively with local search-based and simulated annealing-based cost minimization approaches. The genetic algorithm-based technique is shown to perform very well in terms of robustness to noise, rate of convergence and quality of the final edge image.

Traditional, semantic, and hypersemantic approaches to data modeling

An overview is given of past present data-modeling trends, and future directions are identified. The three traditional and commonly used data models that gained wide acceptance in the late 1960s and early 1970s and are used extensively today, namely the relational, hierarchical, and network models, are reviewed. Semantic data models that attempt to enhance the representation of operational information by capturing more of the meaning about data values and relationships are described. Enhancements to semantic data models that characterize hypersemantic data models and emphasize capturing inferential relationships are discussed.<<ETX>>

NED-2: an agent-based decision support system for forest ecosystem management

Abstract Decision making for forest ecosystem management can include the use of a wide variety of modeling tools. These tools include vegetation growth models, wildlife models, silvicultural models, GIS, and visualization tools. NED-2 is a robust, intelligent, goal-driven decision support system that integrates tools in each of these categories. NED-2 uses a blackboard architecture and a set of semi-autonomous agents to manage these tools for the user. The blackboard integrates a Microsoft Access database and Prolog clauses, and the agents are implemented in Prolog. A graphical user interface written in Visual C++ provides powerful inventory analysis tools, dialogs for selecting timber, water, ecological, wildlife, and visual goals, and dialogs for defining treatments and building prescriptive management plans. Users can simulate management plans and perform goal analysis on different views of the management unit, where a view is determined by a management plan and a point in time. Prolog agents use growth and yield models to simulate management plans, perform goal analyses on user-specified views of the management unit, display results of plan simulation using GIS tools, and generate hypertext documents containing the results of such analysis. Individual agents use metaknowledge to set up and run external simulation models, to load rule-based models and perform inference, to set up and execute external GIS and visualization systems, and to generate hypertext reports as needed, relieving the user from performing all these tasks.

A web-based expert system for gypsy moth risk assessment

The gypsy moth is one of North Americas most devastating exotic forest pests because it can cause the loss of valuable oak species, degraded aesthetics, loss of wildlife habitat, and detrimental effects on watersheds. Due to the increasingly wide infestation of the gypsy moth, it is important to develop decision aids that help assess the risks of this pest to our forests. Expert systems are a type of decision aid that could be applied to the area of risk assessment. We have developed the Gypsy Moth Expert System to estimate the risk that a forest stand faces from the gypsy moth based on the composition, structure, and management objectives of a particular forest. Risk assessment in this context is developed from forest susceptibility to infestation, vulnerability to damage caused by an infestation, and the hazard that management objectives for a forest may be affected if damage occurs. The system uses a straightforward set of if-then rules to classify risk. The development of a web-based expert system presented significant challenges to maintaining remote user processing integrity.

Using DCOM to support interoperability in forest ecosystem management decision support systems

Forest ecosystems exhibit complex dynamics over time and space. Management of forest ecosystems involves the need to forecast future states of complex systems that are often undergoing structural changes. This in turn requires integration of quantitative science and engineering components with socio-political, regulatory, and economic considerations. The amount of data, information and knowledge involved in the management process is often overwhelming. Integrated decision support systems may help managers make consistently good decisions concerning forest ecosystem management. Integrating computer systems using a system-specific or custom approach has many disadvantages. We compare a variety of current approaches, suggest characteristics that an approach should have, and propose that the Distributed Component Object Model is an approach that is very suitable for forest ecosystem decision support system integration.

A mobile robot for corridor navigation: a multi-agent approach

This project focuses on building an autonomous vehicle as the test bed for the future development of an intelligent wheelchair, by proposing a framework for designing and implementing a mobile robot control program that is easily expandable and portable to other robotic platforms. Using a robot equipped with a minimal set of sensors such as a camera and infrared sensors, our multi-agent based control system is built to tackle various problems encountered during corridor navigation. The control system consists of four agents: an agent responsible for handling sensor inputs, an agent which identifies a corridor using machine vision techniques, an agent which avoids collisions by applying fuzzy logic decision making to proximity data, and an agent responsible for locomotion. In the experiments, the robots performance demonstrates the feasibility of a multi-agent approach.

Extending decision support systems: the integration of data, knowledge, and model management

The proliferation of desktop computing has once again rekindled the interest in making computerized tools available to managers and other decision makers. This paper elaborates on a model that integrates data, knowledge, and model management and shows how decision support systems (DSSs) can be extended to support managers in a truly novel way. The model, the Knowledge/Data Model (KDM), is explained and the significance of its applicability to the management of data, knowledge, and models is illustrated through several examples. KDM continues to evolve and is being applied to domains from computer chip design to production and inventory management systems.

Hyper-semantic data modeling

Abstract This paper describes a new area of data modeling, a model in this new area, and the schema specification language for the model. The Knowledge/Data Model captures both knowledge semantics, as specified in Knowledge Based Systems, and data semantics, as represented by Semantic Data Models. The Knowledge/Data Model is an instance of a new class of models, called hyper-semantic data models, which facilitate the incorporation of knowledge in the form of heuristics, uncertainty, constraints and other Artificial Intelligence concepts, together with object-oriented concepts found in Semantic Data Models. The unified knowledge/data modeling features are provided via the constructs of the Knowledge/Data Language.

A genetic algorithm to refine input data selection for air temperature prediction using artificial neural networks

The accurate prediction of air temperature is important in many areas of decision-making including agricultural management, transportation and energy management. Previous research has focused on the development of artificial neural network (ANN) models to predict air temperature from one to twelve hours in advance. The inputs to these models included a constant duration of prior data with a fixed resolution for all environmental variables for all prediction horizons. The overall goal of this research was to develop more accurate ANN models that could predict air temperature for each prediction horizon. The specific objective was to determine if the ANN model accuracy could be improved by applying a genetic algorithm (GA) for each prediction horizon to determine the preferred duration and resolution of input prior data for each environmental variable. The ANN models created based on this GA based approach provided smaller errors than the models created based on the existing constant duration and fixed data resolution approach for all twelve prediction horizons. Except for a few cases, the GA generally included a longer duration for prior air temperature data and shorter durations for other environmental variables. The mean absolute errors (MAEs) for the evaluation input patterns of the one-, four-, eight-, and twelve-hour prediction models that were based on this GA approach were 0.564^oC, 1.264^oC, 1.766^oC and 2.018^oC, respectively. These MAEs were improvements of 3.98%, 4.59%, 2.55% and 1.70% compared to the models that were created based on the existing approach for the same corresponding prediction horizons. Thus, the GA based approach to determine the duration and resolution of prior input data resulted in more accurate ANN models than the existing ones for air temperature prediction. Future work could examine the effects of various GA and fitness evaluation parameters that were part of the approach used in this study.