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.

Using Structural and Functional Information in Diagnostic Design

We wish to design a diagnostic for a device from knowledge of its structure and function. the diagnostic should achieve both coverage of the faults that can occur in the device, and should strive to achieve specificity in its diagnosis when it detects a fault. A system is described that uses a simple model of hardware structure and function, representing the device in terms of its internal primitive functions and connections. The system designs a diagnostic in three steps. First, an extension of path sensitization is used to design a test for each of the connections in teh device. Next, the resulting tests are improved by increasing their specificity. Finally the tests are ordered so that each relies on the fewest possible connections. We describe an implementation of this system and show examples of the results for some simple devices.

SIR: A Computer Program for Semantic Information Retrieval

SIR is a computer system, programmed in the LISP language, which accepts information and answers questions expressed in a restricted form of English. This system demonstrates what can reasonably be called an ability to "understand" semantic information. SIR's semantic and deductive ability is based on the construction of an internal model, which uses word associations and property lists, for the relational information normally conveyed in conversational statements. A format-matching procedure extracts semantic content from English sentences. If an input sentence is declarative, the system adds appropriate information to the model. If an input sentence is a question, the system searches the model until it either finds the answer or determines why it cannot find the answer. In all cases SIR reports its conclusions. The system has some capacity to recognize exceptions to general rules, resolve certain semantic ambiguities, and modify its model structure in order to save computer memory space. Judging from its conversational ability, SIR, is a first step toward intelligent man-machine communication. The author proposes a next step by describing how to construct a more general system which is less complex and yet more powerful than SIR. This proposed system contains a generalized version of the SIR model, a formal logical system called SIR1, and a computer program for testing the truth of SIR1 statements with respect to the generalized model by using partial proof procedures in the predicate calculus. The thesis also describes the formal properties of SIR1 and how they relate to the logical structure of SIR.

Theory and practice in strategic planning for health information systems

Bradshaw, K. & Urquhart, C. (2005). Theory and practice in strategic planning for health information systems. In: D. Wainwright (Ed.), UK Academy for Information Systems 10th conference 2005, 22-24 March 2005 (CD-ROM). Newcastle upon Tyne: Northumbria University.

Origin of avian genome size and structure in non-avian dinosaurs

Chris L. Organ, Andrew M. Shedlock, Andrew Meade, Mark Pagel and Scott V. Edwards (2007). Origin of avian genome size and structure in non-avian dinosaurs. Nature, 46(7132), 180-184. RAE2008

Integration of relational and hierarchical network information for protein function prediction

BACKGROUND:In the current climate of high-throughput computational biology, the inference of a protein's function from related measurements, such as protein-protein interaction relations, has become a canonical task. Most existing technologies pursue this task as a classification problem, on a term-by-term basis, for each term in a database, such as the Gene Ontology (GO) database, a popular rigorous vocabulary for biological functions. However, ontology structures are essentially hierarchies, with certain top to bottom annotation rules which protein function predictions should in principle follow. Currently, the most common approach to imposing these hierarchical constraints on network-based classifiers is through the use of transitive closure to predictions.RESULTS:We propose a probabilistic framework to integrate information in relational data, in the form of a protein-protein interaction network, and a hierarchically structured database of terms, in the form of the GO database, for the purpose of protein function prediction. At the heart of our framework is a factorization of local neighborhood information in the protein-protein interaction network across successive ancestral terms in the GO hierarchy. We introduce a classifier within this framework, with computationally efficient implementation, that produces GO-term predictions that naturally obey a hierarchical 'true-path' consistency from root to leaves, without the need for further post-processing.CONCLUSION:A cross-validation study, using data from the yeast Saccharomyces cerevisiae, shows our method offers substantial improvements over both standard 'guilt-by-association' (i.e., Nearest-Neighbor) and more refined Markov random field methods, whether in their original form or when post-processed to artificially impose 'true-path' consistency. Further analysis of the results indicates that these improvements are associated with increased predictive capabilities (i.e., increased positive predictive value), and that this increase is consistent uniformly with GO-term depth. Additional in silico validation on a collection of new annotations recently added to GO confirms the advantages suggested by the cross-validation study. Taken as a whole, our results show that a hierarchical approach to network-based protein function prediction, that exploits the ontological structure of protein annotation databases in a principled manner, can offer substantial advantages over the successive application of 'flat' network-based methods.

An Information Theoretic Framework for Field Monitoring Using Autonomously Mobile Sensors

We consider a mobile sensor network monitoring a spatio-temporal field. Given limited cache sizes at the sensor nodes, the goal is to develop a distributed cache management algorithm to efficiently answer queries with a known probability distribution over the spatial dimension. First, we propose a novel distributed information theoretic approach in which the nodes locally update their caches based on full knowledge of the space-time distribution of the monitored phenomenon. At each time instant, local decisions are made at the mobile nodes concerning which samples to keep and whether or not a new sample should be acquired at the current location. These decisions account for minimizing an entropic utility function that captures the average amount of uncertainty in queries given the probability distribution of query locations. Second, we propose a different correlation-based technique, which only requires knowledge of the second-order statistics, thus relaxing the stringent constraint of having a priori knowledge of the query distribution, while significantly reducing the computational overhead. It is shown that the proposed approaches considerably improve the average field estimation error by maintaining efficient cache content. It is further shown that the correlation-based technique is robust to model mismatch in case of imperfect knowledge of the underlying generative correlation structure.

Phase estimation receiver for full-field detection: a novel receiver structure for electronic dispersion compensation of metropolitan area networks

The development of ultra high speed (~20 Gsamples/s) analogue to digital converters (ADCs), and the delayed deployment of 40 Gbit/s transmission due to the economic downturn, has stimulated the investigation of digital signal processing (DSP) techniques for compensation of optical transmission impairments. In the future, DSP will offer an entire suite of tools to compensate for optical impairments and facilitate the use of advanced modulation formats. Chromatic dispersion is a very significant impairment for high speed optical transmission. This thesis investigates a novel electronic method of dispersion compensation which allows for cost-effective accurate detection of the amplitude and phase of the optical field into the radio frequency domain. The first electronic dispersion compensation (EDC) schemes accessed only the amplitude information using square law detection and achieved an increase in transmission distances. This thesis presents a method by using a frequency sensitive filter to estimate the phase of the received optical field and, in conjunction with the amplitude information, the entire field can be digitised using ADCs. This allows DSP technologies to take the next step in optical communications without requiring complex coherent detection. This is of particular of interest in metropolitan area networks. The full-field receiver investigated requires only an additional asymmetrical Mach-Zehnder interferometer and balanced photodiode to achieve a 50% increase in EDC reach compared to amplitude only detection.

An examination of the underlying trait structure of psychopathy: applications for screening and measurement

This thesis seeks to clarify the faceted organisation of psychopathy with a view to developing a comprehensive protocol for the assessment of core psychopathic personality traits. The framework developed will, as best as possible, be free of sample bias. The Self-and Informant-report Deviant Personality Screen (DPS) is introduced and a series of empirical studies are conducted to examine the psychometric properties and construct validity of these measures in general and offender populations. Findings from these studies provide strong support for the utility of the DPS scales for the appraisal of psychopathy across diverse population samples. In addition to this, the utility of cognitive based performance measures for the assessment of emotional deficits in psychopathy is evaluated. Results from this study suggest limited correspondence between these measurement techniques and self-report psychopathy measures. Finally, research conducted on offenders suggests that information obtained from DPS reports may be useful within a broad framework of risk assessment. Further empirical and theoretical implications of the research are discussed.

Reconstruction of gross avian genome structure, organization and evolution suggests that the chicken lineage most closely resembles the dinosaur avian ancestor.

BACKGROUND: The availability of multiple avian genome sequence assemblies greatly improves our ability to define overall genome organization and reconstruct evolutionary changes. In birds, this has previously been impeded by a near intractable karyotype and relied almost exclusively on comparative molecular cytogenetics of only the largest chromosomes. Here, novel whole genome sequence information from 21 avian genome sequences (most newly assembled) made available on an interactive browser (Evolution Highway) was analyzed. RESULTS: Focusing on the six best-assembled genomes allowed us to assemble a putative karyotype of the dinosaur ancestor for each chromosome. Reconstructing evolutionary events that led to each species' genome organization, we determined that the fastest rate of change occurred in the zebra finch and budgerigar, consistent with rapid speciation events in the Passeriformes and Psittaciformes. Intra- and interchromosomal changes were explained most parsimoniously by a series of inversions and translocations respectively, with breakpoint reuse being commonplace. Analyzing chicken and zebra finch, we found little evidence to support the hypothesis of an association of evolutionary breakpoint regions with recombination hotspots but some evidence to support the hypothesis that microchromosomes largely represent conserved blocks of synteny in the majority of the 21 species analyzed. All but one species showed the expected number of microchromosomal rearrangements predicted by the haploid chromosome count. Ostrich, however, appeared to retain an overall karyotype structure of 2n=80 despite undergoing a large number (26) of hitherto un-described interchromosomal changes. CONCLUSIONS: Results suggest that mechanisms exist to preserve a static overall avian karyotype/genomic structure, including the microchromosomes, with widespread interchromosomal change occurring rarely (e.g., in ostrich and budgerigar lineages). Of the species analyzed, the chicken lineage appeared to have undergone the fewest changes compared to the dinosaur ancestor.

Computational Methods for RNA Structure Validation and Improvement.

With increasing recognition of the roles RNA molecules and RNA/protein complexes play in an unexpected variety of biological processes, understanding of RNA structure-function relationships is of high current importance. To make clean biological interpretations from three-dimensional structures, it is imperative to have high-quality, accurate RNA crystal structures available, and the community has thoroughly embraced that goal. However, due to the many degrees of freedom inherent in RNA structure (especially for the backbone), it is a significant challenge to succeed in building accurate experimental models for RNA structures. This chapter describes the tools and techniques our research group and our collaborators have developed over the years to help RNA structural biologists both evaluate and achieve better accuracy. Expert analysis of large, high-resolution, quality-conscious RNA datasets provides the fundamental information that enables automated methods for robust and efficient error diagnosis in validating RNA structures at all resolutions. The even more crucial goal of correcting the diagnosed outliers has steadily developed toward highly effective, computationally based techniques. Automation enables solving complex issues in large RNA structures, but cannot circumvent the need for thoughtful examination of local details, and so we also provide some guidance for interpreting and acting on the results of current structure validation for RNA.

Temporal structure and enablement representation for mutual wills: A Petri net approach

Those temporal formalisms that are sporadically found nowadays in the literature of AI & Law are based on temporal logic. We claim a revived role for another major class of temporal representation: Petri nets. This formalism, popular in computing from the 1970s, had its potential recognized on occasion in the literature of legal computing as well, but apparently the discipline has lost sight of it, and its practitioners on average need be tutored into this kind of representation. Asynchronous, concurrent processes—for which the approach is well‐suited—are found in the legal domain, in disparate contexts. We develop an example for Mutual Wills.

Simultaneous Determination Of Fluctuating Age Structure And Mortality From Field Data

The use of a Leslie matrix for analysis of a population normally implies that the age structure of the population is known. However, this restriction can be overcome if the population can be partitioned into recognisably different stages, and some information on stage duration and fecundity is available, in which case the age structure may be determined by the analysis itself. As an example of this approach we consider the estimation of the mortality rate applying to a population from a sequence of observed stage frequency vectors. The technique does not require that the population has attained a stable age structure nor that distinct cohorts can be recognised.

Effects of increases in temperature and nutrients on phytoplankton community structure and photosynthesis in the Western English Channel

Anthropogenic climate change is exerting pressures on coastal ecosystems through increases in temperature, precipitation and ocean acidification. Phytoplankton community structure and photo-physiology are therefore adapting to these conditions. Changes in phytoplankton biomass and photosynthesis in relation to temperature and nutrient concentrations were assessed using a 14 year dataset from a coastal station in the Western English Channel (WEC). Dinoflagellate and coccolithophorid biomass exhibited a positive correlation with temperature, reaching the highest biomass at between 15 and 17°C. Diatoms showed a negative correlation with temperature, with highest biomass at 10°C. Chlorophyll a (chl a) normalised light-saturated photosynthetic rates (PBm) exhibited a hyperbolic response to increasing temperature, with an initial linear increase from 8 to 11°C, and reaching a plateau from 12°C. There was however no significant positive correlation between nutrients and phytoplankton biomass or PBm, which reflects the lag time between nutrient input and phytoplankton growth at this coastal site. The major phytoplankton groups that occurred at this site occupied distinct thermal niches, which in turn modified PBm. Increasing temperature, and higher water column stratification, was major factors in the initiation of dinoflagellates blooms at this site. Dinoflagellates blooms during summer also co-varied with silicate concentration, and acted as a tracer of dissolved inorganic nitrogen and phosphate from river run-off, which were subsequently reduced during these blooms. The data implies that increasing temperature and high river runoff during summer, will promote dinoflaglellates blooms in the WEC.

Can benthic community structure be used to predict the process of bioturbation in real ecosystems

Disentangling the roles of environmental change and natural environmental variability on biologically mediated ecosystem processes is paramount to predict future marine ecosystem functioning. Bioturbation, the biogenic mixing of sediments, has a regulating role in marine biogeochemical processes. However, our understanding of bioturbation as a community level process and of its environmental drivers is still limited by loose use of terminology, and a lack of consensus about what bioturbation is. To help resolve these challenges, this empirical study investigated the links between four different attributes of bioturbation (bioturbation depth, activity and distance, and biodiffusive transport); the ability of an index of bioturbation (BPc) to predict each of them; and their relation to seasonality, in a shallow coastal system – the Western Channel Observatory, UK. Bioturbation distance depended on changes in benthic community structure, while the other three attributes were more directly influenced by seasonality in food availability. In parallel, BPc successfully predicted bioturbation distance but not the other attributes of bioturbation. This study therefore highlights that community bioturbation results from this combination of processes responding to environmental variability at different time-scales. However, community level measurements of bioturbation across environmental variability are still scarce, and BPc is calculated using commonly available data on benthic community structure and the functional classification of invertebrates. Therefore, BPc could be used to support the growth of landscape scale bioturbation research, but future uses of the index need to consider which bioturbation attributes the index actually predicts. As BPc predicts bioturbation distance, estimated here using a random-walk model applicable to community settings, studies using either of the metrics should be directly comparable and contribute to a more integrated future for bioturbation research.