Internet Fish

I have invented "Internet Fish," a novel class of resource-discovery tools designed to help users extract useful information from the Internet. Internet Fish (IFish) are semi-autonomous, persistent information brokers; users deploy individual IFish to gather and refine information related to a particular topic. An IFish will initiate research, continue to discover new sources of information, and keep tabs on new developments in that topic. As part of the information-gathering process the user interacts with his IFish to find out what it has learned, answer questions it has posed, and make suggestions for guidance. Internet Fish differ from other Internet resource discovery systems in that they are persistent, personal and dynamic. As part of the information-gathering process IFish conduct extended, long-term conversations with users as they explore. They incorporate deep structural knowledge of the organization and services of the net, and are also capable of on-the-fly reconfiguration, modification and expansion. Human users may dynamically change the IFish in response to changes in the environment, or IFish may initiate such changes itself. IFish maintain internal state, including models of its own structure, behavior, information environment and its user; these models permit an IFish to perform meta-level reasoning about its own structure. To facilitate rapid assembly of particular IFish I have created the Internet Fish Construction Kit. This system provides enabling technology for the entire class of Internet Fish tools; it facilitates both creation of new IFish as well as additions of new capabilities to existing ones. The Construction Kit includes a collection of encapsulated heuristic knowledge modules that may be combined in mix-and-match fashion to create a particular IFish; interfaces to new services written with the Construction Kit may be immediately added to "live" IFish. Using the Construction Kit I have created a demonstration IFish specialized for finding World-Wide Web documents related to a given group of documents. This "Finder" IFish includes heuristics that describe how to interact with the Web in general, explain how to take advantage of various public indexes and classification schemes, and provide a method for discovering similarity relationships among documents.

Learning World Models in Environments with Manifest Causal Structure

This thesis examines the problem of an autonomous agent learning a causal world model of its environment. Previous approaches to learning causal world models have concentrated on environments that are too "easy" (deterministic finite state machines) or too "hard" (containing much hidden state). We describe a new domain --- environments with manifest causal structure --- for learning. In such environments the agent has an abundance of perceptions of its environment. Specifically, it perceives almost all the relevant information it needs to understand the environment. Many environments of interest have manifest causal structure and we show that an agent can learn the manifest aspects of these environments quickly using straightforward learning techniques. We present a new algorithm to learn a rule-based causal world model from observations in the environment. The learning algorithm includes (1) a low level rule-learning algorithm that converges on a good set of specific rules, (2) a concept learning algorithm that learns concepts by finding completely correlated perceptions, and (3) an algorithm that learns general rules. In addition this thesis examines the problem of finding a good expert from a sequence of experts. Each expert has an "error rate"; we wish to find an expert with a low error rate. However, each expert's error rate and the distribution of error rates are unknown. A new expert-finding algorithm is presented and an upper bound on the expected error rate of the expert is derived.

Model Selection for Solving Kinematics Problems

There has been much interest in the area of model-based reasoning within the Artificial Intelligence community, particularly in its application to diagnosis and troubleshooting. The core issue in this thesis, simply put, is, model-based reasoning is fine, but whence the model? Where do the models come from? How do we know we have the right models? What does the right model mean anyway? Our work has three major components. The first component deals with how we determine whether a piece of information is relevant to solving a problem. We have three ways of determining relevance: derivational, situational and an order-of-magnitude reasoning process. The second component deals with the defining and building of models for solving problems. We identify these models, determine what we need to know about them, and importantly, determine when they are appropriate. Currently, the system has a collection of four basic models and two hybrid models. This collection of models has been successfully tested on a set of fifteen simple kinematics problems. The third major component of our work deals with how the models are selected.

Rodent models for resolving extremes of exercise and health

The extremes of exercise capacity and health are considered a complex interplay between genes and the environment. In general, the study of animal models has proven critical for deep mechanistic exploration that provides guidance for focused and hypothesis driven discovery in humans. Hypotheses underlying molecular mechanisms of disease, and gene/tissue function can be tested in rodents in order to generate sufficient evidence to resolve and progress our understanding of human biology. Here we provide examples of three alternative uses of rodent models that have been applied successfully to advance knowledge that bridges our understanding of the connection between exercise capacity and health status. Firstly we review the strong association between exercise capacity and all-cause morbidity and mortality in humans through artificial selection on low and high exercise performance in the rat and the consequent generation of the "energy transfer hypothesis". Secondly we review specific transgenic and knock-out mouse models that replicate the human disease condition and performance. This includes human glycogen storage diseases (McArdle and Pompe) and α-actinin-3 deficiency. Together these rodent models provide an overview of the advancements of molecular knowledge required for clinical translation. Continued study of these models in conjunction with human association studies will be critical to resolving the complex gene-environment interplay linking exercise capacity, health, and disease.

An artificial immune system for data analysis

Timmis J and Neal M J. An artificial immune system for data analysis. In Proceedings of 3rd international workshop on information processing in cells and tissues (IPCAT), Indianapolis, U.S.A., 1999.

Qualitative Circuit Models in Failure Analysis Reasoning

Lee M.H., Qualitative Circuit Models in Failure Analysis Reasoning, AI Journal. vol 111, pp239-276.1999.

Automating Mechanical FMEA Using Functional Models

Hughes, N., Chou E., Price, C. J. Lee M. H.(1999). Automating Mechanical FMEA Using Functional Models, Proceedings 12th Int. Florida AI Research Soc. Conf. (FLAIRS-99), AAAI Press, May 1999, pp. 394-398.

The Influence of Epistemology on the Design of Artificial Agents

Lee, M. H., Lacey, N. J. (2003). The Influence of Epistemology on the Design of Artificial Agents. Minds and Machines, 13 (3), 367-395

Learning Qualitative Metabolic Models

Coghill, G. M., Garrett, S. M. and King, R. D. (2004) Learning Qualitative Metabolic Models. European Conference on Artificial Intelligence (ECAI'04)

An Empirical Study of the Use of Relevance Information in Inductive Logic Programming

Srinivasan, A., King, R. D. and Bain, M.E. (2003) An Empirical Study of the Use of Relevance Information in Inductive Logic Programming. Journal of Machine Learning Research. 4(Jul):369-383

Generating complex connectivity structures for large-scale neural models

Martin Huelse: Generating complex connectivity structures for large-scale neural models. In: V. Kurkova, R. Neruda, and J. Koutnik (Eds.): ICANN 2008, Part II, LNCS 5164, pp. 849?858, 2008. Sponsorship: EPSRC

Simultaneous Learning of Nonlinear Manifold and Dynamical Models for High-dimensional Time Series

The goal of this work is to learn a parsimonious and informative representation for high-dimensional time series. Conceptually, this comprises two distinct yet tightly coupled tasks: learning a low-dimensional manifold and modeling the dynamical process. These two tasks have a complementary relationship as the temporal constraints provide valuable neighborhood information for dimensionality reduction and conversely, the low-dimensional space allows dynamics to be learnt efficiently. Solving these two tasks simultaneously allows important information to be exchanged mutually. If nonlinear models are required to capture the rich complexity of time series, then the learning problem becomes harder as the nonlinearities in both tasks are coupled. The proposed solution approximates the nonlinear manifold and dynamics using piecewise linear models. The interactions among the linear models are captured in a graphical model. By exploiting the model structure, efficient inference and learning algorithms are obtained without oversimplifying the model of the underlying dynamical process. Evaluation of the proposed framework with competing approaches is conducted in three sets of experiments: dimensionality reduction and reconstruction using synthetic time series, video synthesis using a dynamic texture database, and human motion synthesis, classification and tracking on a benchmark data set. In all experiments, the proposed approach provides superior performance.

Simultaneous Learning of Non-Linear Manifold and Dynamical Models for High-Dimensional Time Series

The goal of this work is to learn a parsimonious and informative representation for high-dimensional time series. Conceptually, this comprises two distinct yet tightly coupled tasks: learning a low-dimensional manifold and modeling the dynamical process. These two tasks have a complementary relationship as the temporal constraints provide valuable neighborhood information for dimensionality reduction and conversely, the low-dimensional space allows dynamics to be learnt efficiently. Solving these two tasks simultaneously allows important information to be exchanged mutually. If nonlinear models are required to capture the rich complexity of time series, then the learning problem becomes harder as the nonlinearities in both tasks are coupled. The proposed solution approximates the nonlinear manifold and dynamics using piecewise linear models. The interactions among the linear models are captured in a graphical model. The model structure setup and parameter learning are done using a variational Bayesian approach, which enables automatic Bayesian model structure selection, hence solving the problem of over-fitting. By exploiting the model structure, efficient inference and learning algorithms are obtained without oversimplifying the model of the underlying dynamical process. Evaluation of the proposed framework with competing approaches is conducted in three sets of experiments: dimensionality reduction and reconstruction using synthetic time series, video synthesis using a dynamic texture database, and human motion synthesis, classification and tracking on a benchmark data set. In all experiments, the proposed approach provides superior performance.

Modeling Financial Time Series with Artificial Neural Networks

Financial time series convey the decisions and actions of a population of human actors over time. Econometric and regressive models have been developed in the past decades for analyzing these time series. More recently, biologically inspired artificial neural network models have been shown to overcome some of the main challenges of traditional techniques by better exploiting the non-linear, non-stationary, and oscillatory nature of noisy, chaotic human interactions. This review paper explores the options, benefits, and weaknesses of the various forms of artificial neural networks as compared with regression techniques in the field of financial time series analysis.

Specification of optimum holistic building environmental and energy performance information to support informed decision making

Political drivers such as the Kyoto protocol, the EU Energy Performance of Buildings Directive and the Energy end use and Services Directive have been implemented in response to an identified need for a reduction in human related CO2 emissions. Buildings account for a significant portion of global CO2 emissions, approximately 25-30%, and it is widely acknowledged by industry and research organisations that they operate inefficiently. In parallel, unsatisfactory indoor environmental conditions have proven to negatively impact occupant productivity. Legislative drivers and client education are seen as the key motivating factors for an improvement in the holistic environmental and energy performance of a building. A symbiotic relationship exists between building indoor environmental conditions and building energy consumption. However traditional Building Management Systems and Energy Management Systems treat these separately. Conventional performance analysis compares building energy consumption with a previously recorded value or with the consumption of a similar building and does not recognise the fact that all buildings are unique. Therefore what is required is a new framework which incorporates performance comparison against a theoretical building specific ideal benchmark. Traditionally Energy Managers, who work at the operational level of organisations with respect to building performance, do not have access to ideal performance benchmark information and as a result cannot optimally operate buildings. This thesis systematically defines Holistic Environmental and Energy Management and specifies the Scenario Modelling Technique which in turn uses an ideal performance benchmark. The holistic technique uses quantified expressions of building performance and by doing so enables the profiled Energy Manager to visualise his actions and the downstream consequences of his actions in the context of overall building operation. The Ideal Building Framework facilitates the use of this technique by acting as a Building Life Cycle (BLC) data repository through which ideal building performance benchmarks are systematically structured and stored in parallel with actual performance data. The Ideal Building Framework utilises transformed data in the form of the Ideal Set of Performance Objectives and Metrics which are capable of defining the performance of any building at any stage of the BLC. It is proposed that the union of Scenario Models for an individual building would result in a building specific Combination of Performance Metrics which would in turn be stored in the BLC data repository. The Ideal Data Set underpins the Ideal Set of Performance Objectives and Metrics and is the set of measurements required to monitor the performance of the Ideal Building. A Model View describes the unique building specific data relevant to a particular project stakeholder. The energy management data and information exchange requirements that underlie a Model View implementation are detailed and incorporate traditional and proposed energy management. This thesis also specifies the Model View Methodology which complements the Ideal Building Framework. The developed Model View and Rule Set methodology process utilises stakeholder specific rule sets to define stakeholder pertinent environmental and energy performance data. This generic process further enables each stakeholder to define the resolution of data desired. For example, basic, intermediate or detailed. The Model View methodology is applicable for all project stakeholders, each requiring its own customised rule set. Two rule sets are defined in detail, the Energy Manager rule set and the LEED Accreditor rule set. This particular measurement generation process accompanied by defined View would filter and expedite data access for all stakeholders involved in building performance. Information presentation is critical for effective use of the data provided by the Ideal Building Framework and the Energy Management View definition. The specifications for a customised Information Delivery Tool account for the established profile of Energy Managers and best practice user interface design. Components of the developed tool could also be used by Facility Managers working at the tactical and strategic levels of organisations. Informed decision making is made possible through specified decision assistance processes which incorporate the Scenario Modelling and Benchmarking techniques, the Ideal Building Framework, the Energy Manager Model View, the Information Delivery Tool and the established profile of Energy Managers. The Model View and Rule Set Methodology is effectively demonstrated on an appropriate mixed use existing ‘green’ building, the Environmental Research Institute at University College Cork, using the Energy Management and LEED rule sets. Informed Decision Making is also demonstrated using a prototype scenario for the demonstration building.