Modeling chemical durability of high level waste glass for nuclear waste processing using bootstrap aggregated neural networks is studied in this paper. In order to overcome the difficulty in developing detailed mechanistic models, data driven neural network models are developed from experimental data. A key issue in building neural network models is that model generalization capability cannot be guaranteed due to the potential over-fitting problem and the limitation in the training data. In order to enhance model generalization, bootstrap aggregated neural networks are used in this study. Multiple neural network models are developed from bootstrap resampling replications of the original training data and are combined to give the final prediction. Application results show that accurate and reliable models can be developed using bootstrap aggregated neural networks.
Fast track article for IS&T International Symposium on Electronic Imaging 2020: Color Imaging: Displaying, Processing, Hardcopy, and Applications proceedings.
Printer impulsive noise significantly contributes to the overall product sound quality. Printer manufacturers are seeking accurate impulsive metrics to understand and quantify customer perceptions and acceptability criteria. Numerous research papers [Baird, Inter-Noise 2005; Ali and Bray, NoiseCon 2004; etc.] have correlated both IT specific and general impulsive metrics to listening test results. These studies show that impulsive metrics that factor in frequency content have high correlation to user perception. This paper discusses the development and verification of a high frequency spectral content printer impulsive metric. This calculation excludes impulsive amplitude, using only the frequency content as a measure of user acceptability.
The major cause of annoyance from impulsive noise is the impulsive amplitude. Many methods can characterize impulsive amplitude, such as time-averaged measures, peak amplitudes or real-time levels. For example, the “impulsive noise index” described by ISO 11201 or the “impulsive parameter” described in ISO 7779 are time-averaged metrics. Real-time metrics might include peak sound pressure level or peak loudness in the time domain. Impulsive metrics must be evaluated for effectiveness. That is, the metric must agree with the perceived impulsive amplitude. Impulsive metrics must also be practical. They must be insensitive to sample-to-sample variation while still providing the granularity between passing and failing results. This paper focuses on evaluating the sensitivity, measurement repeatability, and result granularity to determine the practicality of various effective impulsive noise metrics.
This paper presents a study of using neural networks to model the viscosity of simulated vitrified highly active waste over a range of temperatures and compositions. Vitrification is the process of incorporating the highly active liquid waste into the glass by chemically changing the structure of the glass for nuclear fuel reprocessing. A methodology is needed to determine how the viscosity will change as a result of a new feed. Feed forward neural networks are used to model the viscosity of new product glasses. The results are very promising, with a Mean Squared Error (MSE) of 1.8x10-4 on the scaled unseen validation data, highlighting the high accuracy of the model. Sensitivity analysis of the developed model provides insight on the impact of composition on viscosity.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.