For CMOS designs in sub 90nm technologies, statistical methods are necessary to accurately estimate circuit SER considering process variations. However, due the lack of quality models, current (SSER) frameworks have not yet achieved satisfactory accuracy. In this work, we present accurate table-based cell based on which a Monte Carlo SSER analysis framework is built. We further propose heuristic customize use quasirandom sequences, successfully speeds up convergence simulation error and...

In this paper, we present DiSH, a simulator for large discrete models of biological signal transduction pathways, capable simulating networks with multi-valued elements in both deterministic and stochastic manner. The incorporates the timing molecular reactions, which are often not synchronized occur random order, it also takes into account difference between slow fast reactions. DiSH allows changes conditions during simulations, combined changes, uses concept delays models, similar to...

In this paper, we present DiSH, a simulator for large discrete models of biological signal transduction pathways, capable simulating networks with multi-valued elements in both deterministic and stochastic manner. The incorporates the timing molecular reactions, which are often not synchronized occur random order, it also takes into account difference between slow fast reactions. DiSH allows changes conditions during simulations, combined changes, uses concept delays models, similar to...

In this paper, we propose a new active learning algorithm in which the learner chooses samples to be queried from unlabeled data points whose attributes are only partially observed. addition, cost-driven decision framework where query either labels or missing attributes. This problem statement addresses common constraint when building large datasets and applying techniques on them, some of (including labels) significantly harder more costly acquire per point. We take novel approach problem,...

We present a two-stage model-based approach for unsupervised query-by-example spoken term detection (STD) without any annotated data. Compared to the prevailing DTW approaches STD task, HMMs used by can better capture signal distributions and time trajectories of speech with more global view archive; matching model states also significantly reduces computational load. The utterances in archive are first offline decoded into acoustic patterns automatically discovered an way from archive. In...

This article re-examines the soft error effect caused by radiation-induced particles beyond deep submicron regime. Considering impact of process variations, voltage pulse widths transient faults are found no longer monotonically diminishing after propagation, as they were formerly. As a result, rates in scaled electronic designs escape traditional static analysis and seriously underestimated. In this we formulate statistical rate (SSER) problem present two frameworks to cope with...

A three-stage recursive approach is proposed to improve the recovered distribution of electric parameters in a well-logging environment. The first stage executed using conventional linear sampling method (LSM) and contrast source inversion (CSI) method. In second stage, background updated better identify target shape, results stage. third made closer two, which improves near boundary. effect noise also simulated.

A range‐dependent dictionary learning method is proposed to retrieve the sound‐speed profile (SSP) in a water body, with whole computational domain decomposed into multiple range‐independent subdomains. constructed by using World Ocean Atlas depth profiles of temperature and salinity. The simulation results verify efficacy retrieving SSP each subdomain.

An iterative approach, based on the linear sampling method (LSM) and contrast source inversion (CSI) method, is proposed to improve recovered images of multiple targets with layered or continuous profile, including shape distribution electric properties. The difficulties in dealing large high are partly overcome this approach. Typical studied literatures chosen for simulations comparison.

Summary form only given. Well-logging (WL) techniques have been widelyused for oil exploration, in which antenna arrays are deployed along two boreholes on both sides of the target domain. Usually, recovered distributions permittivity and conductivity roughly conformal to shape, except near boundary. Random ripples permittivityand conductivityare observed inside outside target, more significant

A connected-phoneme hidden Markov model (HMM) is proposed to perform automatic segmentation and labeling. Individual phonetic models are first created by a left-to-right HMM. The large HMM formed grouping all these together. Therefore, each state of this big uniquely represents an English phoneme. not trained the Viterbi algorithm since most probable sequence dose necessarily yield correct Learning vector quantization (LVQ2) used train such that phoneme confusions can be reduced. has two...