[Read before The Royal Statistical Society at a meeting organized by the Research

We report results of experiments on several data sets, in particular: Monk's problems data (see [58]), medical data (lymphography, breast cancer, primary tumor - see [30]) and StatLog's data (see [32]). We compare standard methods for extracting laws from decision tables (see [43], [52]), based on rough set (see [42]) and boolean reasoning (see [8]), with the method based on dynamic reducts and dynamic rules (see [3],[4],[5],[6]). We also compare the results of computer experiments on those data sets obtained by applying our system based on rough set methods with the results on the same data sets obtained with help of several data analysis systems known from literature.

We consider the problem of comparing complex hierarchical models in which the number of parameters is not clearly defined. We follow Dempster in examining the posterior distribution of the log-likelihood under each model, from which we derive measures of fit and complexity (the effective number of parameters). These may be combined into a Deviance Information Criterion (DIC), which is shown to have an approximate decision-theoretic justification. Analytic and asymptotic identities reveal the measure of complexity to be a generalisation of a wide range of previous suggestions, with particular reference to the neural network literature. The contributions of individual observations to fit and complexity can give rise to a diagnostic plot of deviance residuals against leverages. The procedure is illustrated in a number of examples, and throughout it is emphasised that the required quantities are trivial to compute in a Markov chain Monte Carlo analysis, and require no analytic work for new...

Rearranged immunoglobulin variable genes are extensively mutated after stimulation of B lymphocytes by antigen. Mutations are likely generated by an error-prone DNA polymerase, and the mismatch repair pathway may process the mispairs. To examine the role of the MSH2 mismatch repair protein in hypermutation, Msh2�/ � mice were immunized with oxazolone, and B cells were analyzed for mutation in their V�Ox1 light chain genes. The frequency of mutation in the repair-deficient mice was similar to that in Msh2�/ � mice, showing that MSH2-dependent mismatch repair does not cause hypermutation. However, there was a striking bias for mutations to occur at germline G and C nucleotides. The results suggest that the hypermutation pathway frequently mutates G�C pairs, and a MSH2-dependent pathway preferentially corrects mismatches at G and C. Key words: biological sciences • genetics • genes, immunoglobulin • mutation • DNA repair Hypermutation of immunoglobulin variable (V) genes occurs in B lymphocytes after antigen stimulation. Mutations are generated 1,000,000 times more frequently in V genes than in other genes, which implies that the

Traditional approaches to the measurement of performance for CAD algorithms involve the use of sets of so-called “benchmark circuits.” In this paper, we demonstrate that current procedures do not produce results which accuratelycharacterize the behavior of the algorithms under study. Indeed, we show that the apparent advances in algorithms which are documented by traditional benchmarking maywell be due to chance, and not due to anynew properties of the algorithms. As an alternative, we introduce a new methodologyfor the characterization of CAD heuristics which employs wellstudied design of experiments methods. We show through numerous examples how such methods can be applied to evaluate the behavior of heuristics used in BDD variable ordering.

The problem of model selection is relevant to many areas of computer vision. Model selection criteria have been used in the vision literature and many more have been proposed in statistics, but the relative strengths of these criteria have not been analyzed in vision. More importantly, suitable extensions to these criteria must be made to solve problems unique to computer vision. Using the problem of surface reconstruction as our context, we analyze existing criteria using simulations and sensor data, introduce new criteria from statistics, develop novel criteria capable of handling unknown error distributions and outliers, and extend model selection criteria to apply to the surface merging problem. The new and existing model selection criteria and merging rules are tested over a wide range of experimental conditions using both synthetic and sensor data. The new surface merging rules improve upon previous results, and work well even at small step heights (h = 2oe) and crease ...

There is conflicting evidence about the relative roles of division and "recruitment" of responding ceils during the early primary immune response. The cells primarily releasing "19S " hemolysin for sheep erythrocytes may be enumerated using the plaque-forming cell technique developed by Jeme and Nordin (1, 2) and independently by Ingraham (3). An exponential increase in plaque-forming cells occurs during the early primary response; doubling times vary from about 5 to 10 hr depending on the dose of antigen and other experimental variables (4-14). A method using mitotic blocking agents to estimate cell cycle times of antibody-formlng cells, in vivo and in vitro, is presented. Results with this method indicate that the increase in numbers of antibody-forming cells which occurs during the early primary response is due primarily to exponential division of responding cells. Increased responses produced by larger antigen doses or adjuvant appear to result from an increased rate of division of responding cells. These findings are contrary to the view that the observed increase in

common marmosets (Callithrixjacchus) undergo ovulation suppression and exhibit low plasma cortisol levels compared to the dominant, breeding female. To determine whether this cortisol difference is mediated by the differential reproductive consequences of social status, we monitored plasma progesterone and cortisol in 32 adult female marmosets while they were housed in heterosexual pairs, during the first 3 days of heterosexual group formation, and while animals were housed in established social groups. Cortisol levels prior to group formation were significantly higher in females exhibiting cyclic ovulatory activity than in anovulatory females but were not predictive of social status. Subsequently, when animals were housed in established social groups, dominant (cyclic) females had significantly higher cortisol levels than did subordinate (anovulatory) females. Cortisoi levels differed between the pre and postgroup formation conditions only in animals that underwent a corresponding onset or termination of ovulatory cyclicity. Cortisol differences between dominant and subordinate female marmosets therefore appear to be associated with differences in reproductive function rather than with social status per se.

We apply rough set methods and boolean reasoning for knowledge discovery from decision tables. It is often impossible to extract general laws from experimental data by computing first all reducts (Pawlak 1991) of a data table (decision table) and next decision rules from these reducts. We have developed an idea of dynamic reducts as a tool allowing to find relevant reducts for the decision rule generation (Bazan 1994a), (Bazan 1994b), (Bazan 1994c), (Nguyen 1993). Tests on several data tables are showing that the application of dynamic reducts leads to the increasing of the classification quality and/or decreasing of the size of decision rule sets. In this paper we present some statistical arguments showing that the introduced stability coefficients of dynamic reducts are proper measures of their quality. Key words: knowledge discovery, rough sets, decision algorithms, machine learning. 1 INTRODUCTION The aim of the paper is to present a method for extracting laws...

This article describes three long-run monetary facts derived by examining data for 110 countries over a 30-year period, using three definitions of a country’s money supply and two subsamples of countries: (1) Growth rates of the money supply and the general price level are highly correlated for all three money definitions, for the full sample of countries, and for both subsamples. (2) The growth rates of money and real output are not correlated, except for a subsample of countries in the Organisation for Economic Co-operation and Development, where these growth rates are positively correlated. (3) The rate of inflation and the growth rate of real output are essentially uncorrelated. The views expressed herein are those of the authors and not necessarily those of the Federal Reserve Bank of Minneapolis or the Federal Reserve System. The Federal Reserve System was established in 1913 to provide an elastic currency, discount commercial credit, and supervise the banking system in the United States. Congress changed those purposes somewhat with the Employment