Learning first-order probabilistic models with combining rules

Many real-world domains exhibit rich relational structure and stochasticity and motivate the development of models that combine predicate logic with probabilities. These models describe probabilistic influences between attributes of objects that are related to each other through known domain relationships. To keep these models succinct, each such influence is considered independent of others, which is called the assumption of “independence of causal influences” (ICI). In this paper, we describe a language that consists of quantified conditional influence statements and captures most relational probabilistic models based on directed graphs. The influences due to different statements are combined using a set of combining rules such as Noisy-OR. We motivate and introduce multi-level combining rules, where the lower level rules combine the influences due to different ground instances of the same statement, and the upper level rules combine the influences due to different statements. We present algorithms and empirical results for parameter learning in the presence of such combining rules. Specifically, we derive and implement algorithms based on gradient descent and expectation maximization for different combining rules and evaluate them on synthetic data and on a real-world task. The results demonstrate that the algorithms are able to learn both the conditional probability distributions of the influence statements and the parameters of the combining rules.     

MobGeoSen: facilitating personal geosensor data collection and visualization using mobile phones

Mobile sensing and mapping applications are becoming more prevalent because sensing hardware is becoming more portable and more affordable. However, most of the hardware uses small numbers of fixed sensors that report and share multiple sets of environmental data which raises privacy concerns. Instead, these systems can be decentralized and managed by individuals in their public and private spaces. This paper describes a robust system called MobGeoSens which enables individuals to monitor their local environment (e.g. pollution and temperature) and their private spaces (e.g. activities and health) by using mobile phones in their day to day life. The MobGeoSen is a combination of software components that facilitates the phone’s internal sensing devices (e.g. microphone and camera) and external wireless sensors (e.g. data loggers and GPS receivers) for data collection. It also adds a new dimension of spatial localization to the data collection process and provides the user with both textual and spatial cartographic displays. While collecting the data, individuals can interactively add annotations and photos which are automatically added and integrated in the visualization file/log. This makes it easy to visualize the data, photos and annotations on a spatial and temporal visualization tool. In addition, the paper will present ways in which mobile phones can be used as noise sensors using an on-device microphone. Finally, we present our experiences with school children using the above mentioned system to measure their exposure to environmental pollution.

Atomic computing

Technological progress comes from pushing hard at the limitsof what is currently possible, not from merely following trendsothers have set. In computing a good illustration of this principleis the life and work of the 19th century computer pioneer CharlesBabbage (1791-1870), who spent most of his adult life tryingto build a digital computer. Babbage first invented such a machinein 1834. He called it the Analytical Engine.     

Top-down proteomic analysis by MALDI-TOF profiling: Concentration-independent biomarkers

Although numerous protein biomarkers have been correlated with advanced disease states, no new clinical assays have been developed. Goals often anticipate disease-specific protein changes that exceed values among healthy individuals, a property common to acute phase reactants. This review considers somewhat different approaches. It focuses on intact protein isoform ratios that present a biomarker without change in the total concentration of the protein. These will seldom be detected by peptide level analysis or by most antibody-based assays. For example, application of an inexpensive method to large sample groups resulted in observation of several polymorphisms, including the first structural polymorphism of apolipoprotein C1. Isoform distribution of this protein was altered and was eventually linked to increased obesity. Numerous other protein isoforms included C- and N-terminal proteolysis, changes of glycoisoform ratios and certain types of sulfhydryl oxidation. While many of these gave excellent statistical correlation with advanced disease, clinical utility was not apparent. More important may be that protein isoform ratios were very stable in each individual. Diagnosis by longitudinal analysis of the same individual might increase sensitivity of protein biomarkers by 20-fold or more. Protein changes that exceed the range of values found among healthy individuals may be uncommon.     

Practical considerations for high spatial and temporal resolution dynamic transmission electron microscopy

Although recent years have seen significant advances in the spatial resolution possible in the transmission electron microscope (TEM), the temporal resolution of most microscopes is limited to video rate at best. This lack of temporal resolution means that our understanding of dynamic processes in materials is extremely limited. High temporal resolution in the TEM can be achieved, however, by replacing the normal thermionic or field emission source with a photoemission source. In this case the temporal resolution is limited only by the ability to create a short pulse of photoexcited electrons in the source, and this can be as short as a few femtoseconds. The operation of the photo-emission source and the control of the subsequent pulse of electrons (containing as many as 5 x 10(7) electrons) create significant challenges for a standard microscope column that is designed to operate with a single electron in the column at any one time. In this paper, the generation and control of electron pulses in the TEM to obtain a temporal resolution <10(-6)s will be described and the effect of the pulse duration and current density on the spatial resolution of the instrument will be examined. The potential of these levels of temporal and spatial resolution for the study of dynamic materials processes will also be discussed.     

Visual lobe shape and search performance for targets of different difficulty

In quantitative models of visual search it has usually been assumed that visual lobe area shape was sufficiently regular to be approximated by a circle or ellipse. However, the irregularities in visual lobe shapes that have been found in studies involving extensive lobe mapping have suggested that lobe shape may have important implications for visual search performance and for the accuracy of mathematical models used for performance prediction. However, no systematic research on the relationship between the shape aspect of visual lobes and search performance seems to have been carried out and no comparisons of visual lobe shape characteristics under the effect of target difficulty have been reported. The current study was conducted to achieve two major objectives in two experiments. Experiment 1 used two different targets (letter 'O' and letter 'Y') to map the visual lobes of subjects in order to provide a systematic and quantitative comparison of lobe shape characteristics and experiment 2 was to investigate the correlation of visual lobe shape characteristics with visual search time under the effect of target difficulty. The visual lobes of 28 subjects were mapped on 24 imaginary and regularly spaced meridians originating from the centre of the visual field to resemble the full field mapping situation. Five categories of shape indices, viz. roundness, boundary smoothness, symmetry, elongation and shape regularity were investigated. The results of this study demonstrated that the visual lobe shapes of subjects elongate horizontally with medium level of roundness, high levels of boundary smoothness, symmetry and regularity for an easy target (O) against a homogeneous background of 'X's. When a difficult target (Y) was used, the visual lobes of the subjects were still elongated horizontally but to a smaller extent and with a low level of roundness, medium level of boundary smoothness and regularity and a similar high level of symmetry to the easy target. Moreover, significant correlations between shape indices and visual search time were found, suggesting mathematical models for predicting search time should not merely rely on area but also should consider visual lobe shape indices. Finally, a universal mathematical model containing several visual lobe shape indices was developed, which was applicable in the prediction of visual search time for a range of similar search tasks.     

Detecting change-points in Markov chains

Markov chains provide a flexible model for dependent random variables with applications in such disciplines as physics, environmental science and economics. In the applied study of Markov chains, it may be of interest to assess whether the transition probability matrix changes during an observed realization of the process. If such changes occur, it would be of interest to estimate the transitions where the changes take place and the probability transition matrix before and after each change. For the case when the number of changes is known, standard likelihood theory is developed to address this problem. The bootstrap is used to aid in the computation of p-values. When the number of changes is unknown, the AIC and BIC measures are used for model selection. The proposed methods are studied empirically and are applied to example sets of data.     

A taxonomy of clutter reduction for information visualisation

Information visualisation is about gaining insight into data through a visual representation. This data is often multivariate and increasingly, the datasets are very large. To help us explore all this data, numerous visualisation applications, both commercial and research prototypes, have been designed using a variety of techniques and algorithms. Whether they are dedicated to geo-spatial data or skewed hierarchical data, most of the visualisations need to adopt strategies for dealing with overcrowded displays, brought about by too much data to fit in too small a display space. This paper analyses a large number of these clutter reduction methods, classifying them both in terms of how they deal with clutter reduction and more importantly, in terms of the benefits and losses. The aim of the resulting taxonomy is to act as a guide to match techniques to problems where different criteria may have different importance, and more importantly as a means to critique and hence develop existing and new techniques.