8. Subset Selection in Regression (Monographs on Statistics and Applied Probability, no. 40). By A. J. Miller. ISBN 0 412 35380 6. Chapman and Hall, London, 1990. 240 pp. £25.00.

Subset Selection in Regression

Statistical Models in Engineering

When QSAR models are fitted, it is important to validate any fitted model-to check that it is plausible that its predictions will carry over to fresh data not used in the model fitting exercise. There are two standard ways of doing this-using a separate hold-out test sample and the computationally much more burdensome leave-one-out cross-validation in which the entire pool of available compounds is used both to fit the model and to assess its validity. We show by theoretical argument and empiric study of a large QSAR data set that when the available sample size is small-in the dozens or scores rather than the hundreds, holding a portion of it back for testing is wasteful, and that it is much better to use cross-validation, but ensure that this is done properly.

Assessing model fit by cross-validation.

An overview is given of the immobilization of high level radioactive waste (HLW) and surplus materials from a variety of commercial and defence sources employing glass and ceramic hosts. A number of specific host materials are reviewed, including borosilicate and phosphate glasses, glass-ceramics and crystalline ceramics. Topics covered include wasteform processing and manufacture, in addition to wasteform stability, durability and mechanical behaviour. Although, at the present time, borosilicate glass is the generally accepted first generation wasteform for the immobilization of HLW, the emergence of new sources of radioactive materials requiring immobilization has renewed interest in many of the alternative candidates. These include, in particular, titanate, zirconate and phosphate based ceramics, together with iron phosphate based glasses and basaltic glass-ceramics. The relative merits and limitations of each host material are compared and discussed, with particular reference to processing considerations and to current and likely future requirements.

The immobilization of high level radioactive wastes using ceramics and glasses

Response sarfxe methodology (RSM) is a collection of tools developed in the 1950s for the purpose of determining optimum operating conditions in applications in the chemical industry. This article reviews the progrrss of RSM in the general areas of experimental design and analysis and indicates how its role has been affected by advanccs in other fields of applied statistics. Current areas of research in RSM are highlighted. and areas for future research are discussed.

https://www.jstor.org/stable/pdfplus/1268813.pdf

Response surface methodology: 1966–1988

A mixture experiment involves varying the proportions of two or more ingredients, called components of the mixture, and studying the changes that occur in the measured properties (responses) of the...

Mixture Experiment Approaches: Examples, Discussion, and Recommendations

The Statistical Analysis of Compositional Data

The concentration of sulfur in Hanford low-activity waste (LAW) glass melter feed will be maintained below the point where the salt accumulates on the melt surface. The allowable concentrations may range from near zero to over 2.05 wt% (of SO3 on a calcined oxide basis) depending on the composition of the melter feed and processing conditions. If the amount of sulfur exceeds the melt tolerance level, a molten salt will accumulate which may upset melter operations and potentially shorten the useful life of the melter. At the Hanford site, relatively conservative limits have traditionally been placed on sulfur loading in melter feed, which in turn significantly increases the amount of LAW glass that will be produced. Crucible-scale sulfur solubility data and scaled melter sulfur tolerance data have been collected on simulated Hanford waste glasses over the last 15 years. These data were compiled and analyzed. An empirical model was developed to predict the solubility of SO3 in glass based on 253 simulated Hanford LAW glass compositions. This model represents the data well, accounting for over 85% of the variation in data, and was well validated. The model was also found to accurately predict the maximum amount of sulfur in melter feed that did not form a salt layer in 13 scaled melter tests of simulated LAW glasses. The model can be used to help estimate glass volumes and make informed decisions on process options (e.g., scale of supplemental LAW treatment facility, and pretreatment facility performance requirements). The model also gives quantitative estimates of component concentration effects on sulfur solubility. The components that increase sulfur solubility most are Li2O > V2O5 > CaO ≈ P2O5 > Na2O ≈ B2O3 > K2O. The components that decrease sulfur solubility most are Cl > Cr2O3 > Al2O3 > ZrO2 ≈ SnO2 > Others (i.e., the sum of minor components) ≈SiO2. The order of component effects is similar to previous literature data, in most cases.

Toward Understanding the Effect of Low-Activity Waste Glass Composition on Sulfur Solubility

The usual definition of a mixture experiment requires that the response depend only on the proportions of the mixture components and not on the total amount of the mixture. We consider mixture experiments in which the response also depends on the total amount, and we develop mixture-amount models appropriate for such situations. Models in the component amounts are also considered and are shown to be reduced forms of the mixture-amount models. Examples are used to illustrate the development, interpretation, and comparison of the models.

Models for Mixture Experiments When the Response Depends on the Total Amount

Two FORTRAN programs are presented that generate the extreme vertices and the various dimensional centroids (approximate) of an experimental region described by constraints of the general form wher...

Gregory F. Piepel

Papers

Mixture Experiment Approaches: Examples, Discussion, and Recommendations

The Statistical Analysis of Compositional Data

Toward Understanding the Effect of Low-Activity Waste Glass Composition on Sulfur Solubility

Models for Mixture Experiments When the Response Depends on the Total Amount

Programs for Generating Extreme Vertices and Centroids of Linearly Constrained Experimental Regions