Search

PRISM is a probabilistic programming language based on Prolog, augmented with primitives to represent probabilistic choice. It is implemented using a combination of low level support from a modified version of B-Prolog, source level program transformation, and libraries for inference and learning implemented in C. More recently, developers working with functional programming languages have taken the approach of embedding probabilistic primitives into an existing language, with little or no modification to the host language, often by using delimited continuations. Captured continuations represent pieces of the probabilistic program which can be manipulated to achieve a great variety of computational effects useful for inference. In this paper, I will describe an approach based on delimited control operators recently introduced into SWI Prolog. These are used to create a system of nested effect handlers which together implement a core functionality of PRISM—the building of explanation graphs—entirely in Prolog and using an order of magnitude less code. Other declarative programming tools, such as constraint logic programming, are used to implement tools for inference, such as the inside-outside and EM algorithms, lazy best-first explanation search, and MCMC samplers. By embedding the functionality of PRISM into SWI Prolog, users gain access to its rich libraries and development environment. By expressing the functionality of PRISM in a small amount of pure, high-level Prolog, this implementation facilitates further experimentation with the mechanisms of probabilistic logic programming, including new probabilistic modelling features and inference algorithms, such as variational inference in models with real-valued variables.

A basic function of cognition is to detect regularities in sensory input to facilitate the prediction and recognition of future events. It has been proposed that these implicit expectations arise from an internal predictive coding model, based on knowledge acquired through processes such as statistical learning, but it is unclear how different types of statistical information affect listeners' memory for auditory stimuli. We used a combination of behavioral and computational methods to investigate memory for non-linguistic auditory sequences. Participants repeatedly heard tone sequences varying systematically in their information-theoretic properties. Expectedness ratings of tones were collected during three listening sessions, and a recognition memory test was given after each session. Information-theoretic measures of sequential predictability significantly influenced listeners' expectedness ratings, and variations in these properties had a significant impact on memory performance. Predictable sequences yielded increasingly better memory performance with increasing exposure. Computational simulations using a probabilistic model of auditory expectation suggest that listeners dynamically formed a new, and increasingly accurate, implicit cognitive model of the information-theoretic structure of the sequences throughout the experimental session.

We present recent developments on the syntax of Real, a library for interfacing two Prolog systems to the statistical language R. We focus on the changes in Prolog syntax within SWI-Prolog that accommodate greater syntactic integration, enhanced user experience and improved features for web-services. We recount the full syntax and functionality of Real as well as presenting a full application and sister packages which include Prolog code interfacing a number of common and useful tasks that can be delegated to R. We argue that Real is a powerful extension to logic programming, providing access to a popular statistical system that has complementary strengths in areas such as machine learning, statistical inference and visualisation. Furthermore, Real has a central role to play in the uptake of semantic web, computational biology and bioinformatics as application areas for research in logic programming. Presents advances in Real, a bridge from AI language Prolog to R.Tighter syntactic integration creates a powerful programming paradigm.Combining strong knowledge representation with statistical prowess.A web-based musicology application highlights on-line deployments.Sister packages illustrate use of framework and introduces R to logic programmers.

In the transition from traditional to digital musicology, large scale music data are increasingly becoming available which require research methods that work on the collection level and at scale. In the Digital Music Lab (DML) project, a software system has been developed that provides large-scale analysis of music audio with an interactive interface. The DML system includes distributed processing of audio and other music data, remote analysis of copyright-restricted data, logical inference on the extracted information and metadata, and visual web-based interfaces for exploring and querying music collections. A system prototype has been set up in collaboration with the British Library and I Like Music Ltd, which has been used to analyse a diverse corpus of over 250,000 music recordings. In this paper we describe the system requirements, architecture, components, and data sources, explaining their interaction. Use cases and applications with initial evaluations of the proposed system are also reported.

NOTICE: this is the author’s version of a work that was accepted for publication in 'Physical Letters A'. Changes resulting from the publishing process, such as peer review, editing, corrections, structural formatting, and other quality control mechanisms may not be reflected in this document. Changes may have been made to this work since it was submitted for publication. A definitive version was subsequently published in PHYSICAL LETTERS A, 376 (4): 275-281, JAN 2012. DOI:10.1016/j.physleta.2011.10.066.

The variation of dissolved radon (222Rn) levels in water supplies remains of interest because of the radiation-induced public health hazards. A large part of the Lebanese population relies on springs and wells for their drinking water. 222Rn measurements in spring and well water sources were conducted using the E-PERM method at sites ranging from sea level to 1200 m above sea level and across several geologic formations within Lebanon. The dissolved radon concentrations ranged from a low of 0.91 Bq L - 1 in a coastal well source to a high of 49.6 Bq L - 1 for a spring source in a mountainous region. Of the 20 sites sampled, only five had radon levels above 11 Bq L - 1 and these mostly occurred in areas adjacent to well-known geological fault zones. A preliminary national average radon level was determined to be about 11.4 Bq L - 1 . In general, as all determined concentrations were well below the 100 and 146 Bq L - 1 revised reference levels proposed in the European Union and the United States, respectively, it is concluded that there is no reason to believe these water sources pose any radon-related hazard. On the other hand, at locations where water is collected directly from the springhead, it is advisable to have a settling/piping system installed allowing for further radon decay and radon loss into the air to alleviate any possible radon problem.

Recent developments in computational linguistics offer ways to approach the analysis of musical structure by inducing probabilistic models (in the form of grammars) over a corpus of music. These can produce idiomatic sentences from a probabilistic model of the musical language and thus offer explanations of the musical structures they model. This chapter surveys historical and current work in musical analysis using grammars, based on computational linguistic approaches. We outline the theory of probabilistic grammars and illustrate their implementation in Prolog using PRISM. Our experiments on learning the probabilities for simple grammars from pitch sequences in two kinds of symbolic musical corpora are summarized. The results support our claim that probabilistic grammars are a promising framework for computational music analysis, but also indicate that further work is required to establish their superiority over Markov models.

Note onset detection and localization is useful in a number of analysis and indexing techniques for musical signals. The usual way to detect onsets is to look for "transient" regions in the signal, a notion that leads to many definitions: a sudden burst of energy, a change in the short-time spectrum of the signal or in the statistical properties, etc. The goal of this paper is to review, categorize, and compare some of the most commonly used techniques for onset detection, and to present possible enhancements. We discuss methods based on the use of explicitly predefined signal features: the signal's amplitude envelope, spectral magnitudes and phases, time-frequency representations; and methods based on probabilistic signal models: model-based change point detection, surprise signals, etc. Using a choice of test cases, we provide some guidelines for choosing the appropriate method for a given application.

Objectives: To assess whether workers at Lucas Heights Science and Technology Centre (LHSTC) have different levels of mortality from the New South Wales (NSW) and Australian populations. Methods: A retrospective cohort study was undertaken at LHSTC. Data on 7,076 workers employed between 1957-98 were abstracted from personnel, dosimetry, and medical files. Deaths registrations in the cohort were identified to 1998 through electronic linkage of records with NSW and national registers of cancer incidence and mortality. Two inception cohorts were defined as including 4,717 and 3,543 workers in employment between 1972-98 and 1980-98, to examine cancer mortality and all-cause mortality respectively. Results: All-cause mortality was 31% lower than the national rates; all-cancer mortality was 19% below the NSW rate. Of 37 specific cancers and groups of cancers examined, statistically significant excesses relative to NSW rates were observed only for pleural cancer mortality (SMR=21.11; 95% Cl 8.79-50.72). Conclusions: The observed increase in the risk of cancer of the pleura was probably due to unmeasured exposures, given the lack of an established association with radiation exposure and the strong link to asbestos exposure.

This study reports on the regulation and remodeling role of endothelin-1 (ET-1) and its receptor subtypes, ET(A)-Rs/ET(B)-Rs, at the coronary endothelium (CE) and cardiomyocyte (CM) sites. It is carried out in normal and normotensive rats with streptozotocin-induced diabetes mellitus receiving different treatment modalities. Normal rats were divided into two groups, namely a placebo (N) and a losartan-treated (NL), and diabetic rats into four groups receiving placebo (D), insulin-treated (DI), losartan-treated (DL), and insulin/losartan-treated (DIL) respectively. Binding kinetics of ET-1 to ET(A)-Rs/ET(B)-Rs on CE and CMs were assessed in the above groups to try to explain the effect of therapeutic doses of an angiotensin II receptor subtype-1 blocker on the dynamics of this ligand and its receptor in insulin supplemented diabetic animals. Each group was divided into two subgroups: CHAPS-untreated and CHAPS-treated rat hearts perfused with [125I]ET-1 to respectively estimate ET-1 binding affinity (tau = 1/k-n) to its receptor subtype(s) on CE and CMs using mathematical modeling describing a 1:1 reversible binding stoichiometry. Heart perfusion results revealed that insulin treatment significantly decreased tau on CE but not on CMs in diabetic rats. In diabetics treated with losartan, an increase in tau value on CE but not on CMs was noted. Cotreatment of diabetic rats with insulin and losartan normalized tau on CE but decreased it on CMs. Western blot, using snap-frozen heart tissues, revealed increase in ET(A)-R density in all diabetic groups. However, significant decrease in ET(B)-R density was observed in all groups compared to the normal, and was reconfirmed by immunohistochemical analysis. In conclusion, coadministration of insulin and losartan in nonhypertensive animals suffering from diabetes type 1 may offer new cardiac protection benefits by improving coronary blood flow and cardiomyocyte contractility through modulating ET-1 receptor subtypes density and affinity at CE and CM sites.

The output of the sheet metal forming process is subject to much variation. This paper develops a method to measure shape variation in channel forming and relate this back to the corresponding process parameter levels of the manufacturing set-up to create an inverse model. The shape variation in the channels is measured using a modified form of the point distribution model (also known as the active shape model). This means that channels can be represented by a weighting vector of minimal linear dimension that contains all the shape variation information from the average formed channel. The inverse models were created using classifiers that related the weighting vectors to the process parameter levels for the blank holder force (BHF), die radii (DR) and tool gap (TG) of the parameters. Several classifiers were tested: linear, quadratic Gaussian and artificial neural networks. The quadratic Gaussian classifiers were the most accurate and the most consistent type of classifier over all the parameters.

There are many variations within sheet metal forming, some of which are manifest in the final geometry of the formed component. It is important that this geometric variation be quantified and measured for use in a process or quality control system. The contribution of this paper is to propose a novel way of measuring the geometric difference between the desired shape and an actual formed “U” channel. The metric is based upon measuring errors in terms of the significant manufacturing variations. The metric accords with the manually measured errors of the channel set. The shape error metric is then extended to develop a simple empirical, whole-component, springback error measure. The springback error measure combines into one value all the angle springback and side wall curl geometric errors for a single channel. Two trends were observed: combined springback decreases when the blank holder force is increased; and the combined springback marginally decreases when the die radii is increased.

The complexity of the forging process ensures that there is inherent variability in the geometric shape of a forged part. While knowledge of shape error, comparing the desired versus the measured shape, is significant in measuring part quality the question of more interest is what can this error suggest about the forging process set-up? The first contribution of this paper is to develop a shape error metric which identifies geometric shape differences that occur from a desired forged part. This metric is based on the point distribution deformable model developed in pattern recognition research. The second contribution of this paper is to propose an inverse model that identifies changes in process set-up parameter values by analysing the proposed shape error metric. The metric and inverse models are developed using two sets of simulated hot-forged parts created using two different die pairs (simple and ‘M’-shaped die pairs). A neural network is used to classify the shape data into three arbitrarily chosen levels for each parameter and it is accurate to at least 77 per cent in the worst case for the simple die pair data and has an average accuracy of approximately 80 per cent when classifying the more complex ‘M’-shaped die pair data.

This paper examined a number of problems with the derivation of the approximation of the model-IR, including an algebraic error, some inconsistent or untenable assumptions, and an unsupported approximation. These problems make it difficult to judge the significance of the results presented therein.

We describe an information-theoretic approach to the analysis of music and other sequential data, which emphasises the predictive aspects of perception, and the dynamic process of forming and modifying expectations about an unfolding stream of data, characterising these using the tools of information theory: entropies, mutual informations, and related quantities. After reviewing the theoretical foundations, we discuss a few emerging areas of application, including musicological analysis, real-time beat-tracking analysis, and the generation of musical materials as a cognitively-informed compositional aid. © 2012 IEEE.

This paper adapts an Ant Colony intelligent Optimization algorithm (ACO) to four design problems important in Mechatronics applications: (1) Stress- and volume minimization of a helical compression spring subject to a set of geometrical and stress constraints for material volume minimization, (2) tuning the gains of a PID-controller for a second-order plant with a transport lag, (3) design of a planetary gear leg of a walking machine for a desired straight line trajectory and driving actuator torque, and (4) synthesis of four-bar mechanisms. A brief overview of the ACO algorithm is first presented and its adaptation to each design problem is explained and demonstrated by way of examples that have been solved by other optimization techniques for comparison. In all four examples, the ACO algorithm produced better results in terms of cost-function values while satisfying all necessary constraints.Copyright © 2012 by ASME

Most sound scenes result from the superposition of several sources, which can be separately perceived and analyzed by human listeners. Source separation aims to provide machine listeners with similar skills by extracting the sounds of individual sources from a given scene. Existing separation systems operate either by emulating the human auditory system or by inferring the parameters of probabilistic sound models. In this chapter, the authors focus on the latter approach and provide a joint overview of established and recent models, including independent component analysis, local time-frequency models and spectral template-based models. They show that most models are instances of one of the following two general paradigms: linear modeling or variance modeling. They compare the merits of either paradigm and report objective performance figures. They also,conclude by discussing promising combinations of probabilistic priors and inference algorithms that could form the basis of future state-of-the-art systems.

Simultaneous Localization and Mapping (SLAM) is a recursive probabilistic inferencing process for concurrently building a map of a robot’s surroundings and localizing that robot within this map. The ultimate goal of SLAM is to operate anywhere, allowing a robot to navigate autonomously and producing a meaningful purposeful map. Research in SLAM to date has focused on improving the localization part of SLAM, while lagging in the ability to produce useful maps. Indeed, all feature-based SLAM maps are built from either low level features such as points or lines or from artificial beacons; such maps have little use other than to perform SLAM. There are benefits in building maps from real natural objects that are indigenous of the environment for operations such as surveying of remote areas or as a guide for human navigation in dangerous settings. To investigate the potential of SLAM to produce such maps, an Inertial-Visual SLAM system is designed and used here which relies on inertial measurements to predict ego-motion and a digital camera to collect images of natural landmarks about the scene. Experiments conducted on a mobile vehicle show encouraging results and highlight the potential for Vision SLAM to generate meaningful maps which agree with ground truth. The Computer Vision system is capable of recognizing the environment type, of detecting trees within this environment, and recognizing different trees based on clusters of distinctive visual features.