Imprecise Components Research Articles

A Comprehensive Model for Efficient Design Space Exploration of Imprecise Computational Blocks

After almost a decade of research, development of more efficient imprecise computational blocks is still a major concern in imprecise computing domain. There are many instances of the introduced imprecise components of different types, while their main difference is that they propose different precision-cost-performance trade-offs. In this paper, a novel comprehensive model for the imprecise components is introduced, which can be exploited to cover a wide range of precision-cost-performance trade-offs, for different types of imprecise components. The model helps to find the suitable imprecise component based on any desired error criterion. Therefore, the most significant advantage of the proposed model is that it can be simply exploited for design space exploration of different imprecise components to extract the suitable components, with the desired precision-cost-performance trade-off for any specific application. To demonstrate the efficiency of the proposed model, two novel families of Lowest-cost Imprecise Adders (LIAs) and Lowest-cost Imprecise Multipliers (LIMs) are introduced in the paper, which are systematically extracted based on exploration of the design space provided by the proposed model. A wide range of simulation and synthesis results are also presented in the paper to prove the comparable efficiency of the systematically extracted LIA/LIM structures with respect to the most efficient existing human-made imprecise components both individually and in a Multiply-Accumulate application.

RAPID: Approximate Pipelined Soft Multipliers and Dividers for High Throughput and Energy Efficiency

The rapid updates in error-resilient applications along with their quest for high throughput have motivated designing fast approximate functional units for Field-Programmable Gate Arrays (FPGAs). Studies that proposed imprecise functional techniques are posed with three shortcomings: first, most inexact multipliers and dividers are specialized for Application-Specific Integrated Circuit (ASIC) platforms. Second, state-of-the-art (SoA) approximate units are substituted, mostly in a single kernel of a multi-kernel application. Moreover, the end-to-end assessment is adopted on the Quality of Results (QoR), but not on the overall gained performance. Finally, existing imprecise components are not designed to support a pipelined approach, which could boost the operating frequency/throughput of, e.g., division-included applications. In this paper, we propose RAPID, the first pipelined approximate multiplier and divider architecture, customized for FPGAs. The proposed units efficiently utilize 6-input Look-up Tables (6-LUTs) and fast carry chains to implement Mitchell's approximate algorithms. Our novel error-refinement scheme not only has negligible overhead over the baseline Mitchell's approach but also boosts its accuracy to 99.4% for arbitrary size of multiplication and division. Experimental results demonstrate the efficiency of the proposed pipelined and non-pipelined RAPID multipliers and dividers over accurate counterparts. Moreover, the end-to-end evaluations of RAPID, deployed in three multi-kernel applications in the domains of bio-signal processing, image processing, and moving object tracking for Unmanned Air Vehicles (UAV) indicate up to 45% improvements in area, latency, and Area-Delay-Product (ADP), respectively, over accurate kernels, with negligible loss in QoR.

Generalized robust training scheme using genetic algorithm for optical neural networks with imprecise components

One of the pressing issues for optical neural networks (ONNs) is the performance degradation introduced by parameter uncertainties in practical optical components. Hereby, we propose a novel two-step ex situ training scheme to configure phase shifts in a Mach–Zehnder-interferometer-based feedforward ONN, where a stochastic gradient descent algorithm followed by a genetic algorithm considering four types of practical imprecisions is employed. By doing so, the learning process features fast convergence and high computational efficiency, and the trained ONN is robust to varying degrees and types of imprecisions. We investigate the effectiveness of our scheme by using practical machine learning tasks including Iris and MNIST classifications, showing more than 23% accuracy improvement after training and accuracy (90.8% in an imprecise ONN with three hidden layers and 224 tunable thermal-optic phase shifters) comparable to the ideal one (92.0%).

Imprecise Gaussian discriminant classification

Gaussian discriminant analysis is a popular classification model, that in the precise case can produce unreliable predictions in case of high uncertainty (e.g., due to scarce or noisy data). While imprecise probability theory offers a nice theoretical framework to solve such issues, it has not been yet applied to Gaussian discriminant analysis. This work remedies this, by proposing a new Gaussian discriminant analysis based on robust Bayesian analysis and near-ignorance priors. The model delivers cautiouspredictions, in form of set-valued class, in case of limited or imperfect available information. We present and discuss results of experimentation on real and synthetic datasets, where for this latter we corrupt the test instance to see how our approach reacts to non i.i.d. samples. Experiments show that including an imprecise component in the Gaussian discriminant analysis produces reasonably cautious predictions, and that set-valued predictions correspond to instances for which the precise model performs poorly.

A New Approach for the Evaluation of Structural Failure by Credibility Distribution

Aim: To devise an effective structural failure analysis approach under uncertainty. Background: In reliability evaluation, plenty of factors are uncertain, or sometimes, spontaneously represented via linguistic expressions, and as a consequence, the traditionalist appraisal methods cannot capably handle the ambiguity and vagueness that occurs in reliability assessment components. Subsequently, this leads to the problem of tremendous computationally multifaceted and scanty correctness. Objective: To overcome the limitations and to develop efficiency as well as accuracy in structural failure evaluation techniques, an attempt has been made to devise a novel structural reliability assessment method via credibility distribution. Methods: To get rid of the problems of massive computationally difficult and inadequate precision, an algorithm has been devised using credibility sampling. To exhibit the novelty, validity, and applicability of the present approach, some structural failure assessment problems are solved along with a comparison with the existing approach. Results: The proposed method was verified by four examples and applied in structural analysis. It was observed that the present approach is technically sound and efficient; it can overcome all the drawbacks of the existing approach. Moreover, the approach can be executed in any uncertain situation. Conclusion: After evaluation of failure assessment, it is experienced that the increase in the number of simulations leads to better precision. Furthermore, it is encountered that when hybridization problems i.e., representation of imprecise components in the problem of structural failure, are both fuzzy and probabilistic nature, then the failure assessment is attained to be maximum.

Reliability Analysis and Imprecise Component Importance Measure of Redundant Systems of OWTs Based on Component Swapping

Due to the high cost of failures of wind turbines, redundancy designs are commonly applied in wind turbines for improving the reliability and availability of systems. For this reason, replacing failed components with other working components of the same type in redundant systems is becoming an attractive option of maintenance strategies towards more resilient systems. To quantitatively evaluate system’s reliability, this paper focuses on the reliability analysis of redundant systems of offshore wind turbines based on swapping existing components. The survival signature-based component swapping method is introduced to describe the new structure-function of the system upon swapping. Furthermore, the reliability model of redundant systems is established using the fault tree and survival signature. Following this, the influences of component swapping on component reliability importance measure (marginal reliability importance and joint reliability importance) without and with considerations of the imprecision of failure rates are explored. Finally, a 5MW offshore wind turbine is presented to show the applicability of the proposed approach for redundant systems, and the results show that the proposed approach can obtain realistic reliability assessment of redundant systems and considering component swapping can significantly improve system reliability.

Analysis of a Point-of-Care HbA1c Assay: Is It Time for an HbA1c Error Grid?

In an article in Journal of Diabetes Science and Technology, Arnold et al have presented a thorough study of imprecision components for a point-of-care hemoglobin A1c (HbA1c) assay. An interesting and innovative approach is the combination of data from different studies to arrive at a total error estimate. But total error has the oxymoron feature of estimating performance for most (95%) but not all of the results. An HbA1c error grid would provide the severity for results that exceed the 6% requirement. Since this device is intended for Clinical Laboratory Improvement Amendments waived labs and allows for finger-stick samples, monitoring the Food and Drug Administration adverse event database (MAUDE, Manufacturer and User Facility Device Experience) is recommended.

Satisfying uncertain targets using measure generalized Dempster-Shafer belief structures

We provide the basic formalisms associated with the classic Dempster–Shafer belief structure. We note it provides a model of uncertainty that has both a probabilistic uncertainty and imprecise component. We introduce the idea of generalized D–S belief structure, which, while still having an imprecise component has more a general measure based uncertainty. We discuss the idea of a target as a subset of some space X and an uncertain target is a collection of subsets each having a probability of being the true target. We consider the problem that an uncertain variable, whose value is known to be a generalized D–S structure satisfies the target, its value lies in the target set.

Quantifying protein measurands by peptide measurements: where do errors arise?

Clinically actionable quantification of protein biomarkers by mass spectrometry (MS) requires analytical performance in concordance with quality specifications for diagnostic tests. Laboratory-developed tests should, therefore, be validated in accordance with EN ISO 15189:2012 guidelines for medical laboratories to demonstrate competence and traceability along the entire workflow, including the selected standardization strategy and the phases before, during, and after proteolysis. In this study, bias and imprecision of a previously developed MS method for quantification of serum apolipoproteins A-I (Apo A-I) and B (Apo B) were thoroughly validated according to Clinical and Laboratory Standards Institute (CLSI) guidelines EP15-A2 and EP09-A3, using 100 patient sera and either stable-isotope labeled (SIL) peptides or SIL-Apo A-I as internal standard. The systematic overview of error components assigned sample preparation before the first 4 h of proteolysis as major source (∼85%) of within-sample imprecision without external calibration. No improvement in imprecision was observed with the use of SIL-Apo A-I instead of SIL-peptides. On the contrary, when the use of SIL-Apo A-I was combined with external calibration, imprecision improved significantly (from ∼9% to ∼6%) as a result of the normalization for matrix effects on linearity. A between-sample validation of bias in 100 patient sera further supported the presence of matrix effects on digestion completeness and additionally demonstrated specimen-specific biases associated with modified peptide sequences or alterations in protease activity. In conclusion, the presented overview of bias and imprecision components contributes to a better understanding of the sources of errors in MS-based protein quantification and provides valuable recommendations to assess and control analytical quality in concordance with the requirements for clinical use.

Multicomponent noisy signal adaptive instantaneous frequency estimation using components time support information

This study proposes an adaptive method for components instantaneous frequency (IF) estimation of noisy non-stationary multicomponent signals, combined with the components time-support estimation method based on the short-time Rényi entropy (STRE). Components localisation and separation are done using a double-direction component tracking and extraction method presented here, while the IF estimation is done using the adaptive algorithms based on the intersection of confidence intervals (ICI) rule and the relative intersection of confidence intervals (RICI) rule. The results obtained using the ICI and RICI rules are compared for various window types, signal-to-noise ratios and time–frequency distributions, both with and without using the information on components time support. Most of the errors in IF estimation using the ICI and RICI-based methods are caused by imprecise components time-support estimation. The proposed methods combined with the STRE have achieved a significant accuracy improvement in terms of the mean absolute error and the mean squared error, reducing them by up to 73 and 93%, respectively. The method has been applied to real-life signals and proven to be an efficient tool for IF estimation of noisy non-stationary multicomponent signals.

Compositional real-time scheduling framework for periodic reward-based task model

As the size and complexity of embedded software systems increase, compositional real-time scheduling framework is widely accepted as means to build large and complex systems. A compositional real-time scheduling framework proposes to decompose a system into independent subsystems and provides ways to assemble them into a flexible hierarchical real-time scheduling system while guaranteeing the internal real-time requirements of each subsystem. In this paper, we consider the imprecise reward-based periodic task model in compositional scheduling framework. Thus, we introduce the imprecise periodic resource model to characterize the imprecise resource allocations provided by the system to a single component, and the interface model to abstract the imprecise real-time requirements of the component. The schedulability of mandatory parts is also analyzed to meet the minimum requirement of tasks. Finally, we provide a scheduling algorithm to guarantee a certain amount of reward, which makes it feasible to efficiently compose multiple imprecise components.

Systematic monitoring of standardization and harmonization status with commutable EQA-samples—Five year experience from the Netherlands

BackgroundEquivalence of results among laboratories is a major mission for medical laboratories. Monitoring of test equivalence is structurally integrated in the Dutch External Quality Assessment (EQA) scheme since 2005. Commutable poolsera, single donation “spy” sera and biological variance tolerance limits have been introduced in the EQA scheme for evaluation of the degree of test equivalence and its determinants. MethodsIn the annual cycle scheme 24 samples, covering the (patho)physiological measuring range for 17 analytes, are assayed by 220 participating laboratories at biweekly intervals. Test equivalence was evaluated by calculating overall median interlaboratory coefficients of variation (CVs) and its bias and imprecision components. Data from 2005 and 2010 schemes are evaluated to investigate trends in performance and success of standardization efforts. ResultsOverall median interlaboratory CVs in 2010 were mostly better than in 2005. Median interlaboratory CVs became <5% for electrolytes and substrates, and <10% for enzymes. Improvement in median interlaboratory CVs over these five years is mainly explained by improved method standardization, especially for enzymes and creatinine. ConclusionThe Dutch EQA-program proves to be a powerful instrument to evaluate test equivalence. It allows monitoring standardization efforts in a highly effective way and gives insight into remaining standardization potential.

The authors of the article cited above respond:

Krouwer (1) makes a valid point that total error for a measurement procedure must include analytical nonspecificity influences (referred to as “random patient interferences” by Krouwer) as well as mean calibration bias and imprecision components. Fig. 3 in the report from the Laboratory Working Group (LWG) of the National Kidney Disease Education Program showed boundaries for combinations of bias and imprecision that would contribute no more than a 10% change in root mean square error for estimates of glomerular filtration rate …

Monte Carlo simulation techniques for group decision making with incomplete information

In this paper we deal with group decision-making problems where several decision makers elicit their own preferences separately. The decision makers’ preferences are quantified using a decision support system, which admits incomplete information concerning the decision makers’ responses to the questions they are asked. Consequently, each decision maker proposes classes of utility functions and attribute weight intervals for the different attributes. We introduce an approach based on Monte Carlo simulation techniques for aggregating decision maker preferences that could be the starting point for a negotiation process, if necessary. The negotiation process would basically involve the decision maker tightening the imprecise component utilities and weights to output more meaningful results and achieve a consensus alternative. We focus on how attribute weights and the component utilities associated with a consequence are randomly generated in the aggregation process taking into account the decision-makers’ preferences, i.e., their respective attribute weight intervals and classes of utility functions. Finally, an application to the evaluation of intervention strategies for restoring a radionuclide contaminated lake illustrates the usefulness and flexibility of this iterative process.

Monte Carlo Simulation Techniques in a Decision Support System for Group Decision Making

This paper describes a group decision support system based on an additive multi-attribute utility model for identifying a consensus strategy in group decision-making problems where several decision-makers or groups of decision-makers elicit their own preferences separately. On the one hand, the system provides procedures to quantify the DM’s or group of DMs’ preferences separately. This involves assessing the DM’s or group of DMs’ component utilities that represent their preferences regarding the respective possible attribute values and objective weights that represent the relative importance of the criteria. On the other hand, we propose Monte Carlo simulation techniques for identifying a consensus strategy. An iterative process will be carried out, where, after the simulations have been performed, the imprecise component utilities and weights corresponding to the different DMs or groups of DMs are tightened to output more meaningful information in the next simulations to achieve a consensus strategy. Finally, an application to the evaluation of remedial strategies for restoring contaminated aquatic ecosystems illustrates the usefulness and flexibility of this decision support tool.

Methods and apparatus for acoustic transient processing

Method and apparatus for using current-mode analog circuits to compute the correlation between an auditory input signal in the time-frequency domain and a stored binary template. By using massively parallel computation, circuits made of relatively imprecise analog components are able to accurately classify transient events, are competitive with high-performance DSP systems, and operate with much smaller power requirements.

Reverse engineering of biological complexity.

Advanced technologies and biology have extremely different physical implementations, but they are far more alike in systems-level organization than is widely appreciated. Convergent evolution in both domains produces modular architectures that are composed of elaborate hierarchies of protocols and layers of feedback regulation, are driven by demand for robustness to uncertain environments, and use often imprecise components. This complexity may be largely hidden in idealized laboratory settings and in normal operation, becoming conspicuous only when contributing to rare cascading failures. These puzzling and paradoxical features are neither accidental nor artificial, but derive from a deep and necessary interplay between complexity and robustness, modularity, feedback, and fragility. This review describes insights from engineering theory and practice that can shed some light on biological complexity.

The use of imprecise component reliability distributions in reliability calculations : Ian D. Roberts and Andrew E. Samuel. IEEE Transactions on Reliability, 45 (1), 141 (1996)

The use of imprecise component reliability distributions in reliability calculations : Ian D. Roberts and Andrew E. Samuel. IEEE Transactions on Reliability, 45 (1), 141 (1996)

The use of imprecise component reliability distributions in reliability calculations

System reliability prediction can be needed when detailed data concerning each component are unavailable. This prevents data from various sources from being integrated by considering each source as a sample from the same population. For the Weibull distribution, fuzzy arithmetic is used to synthesize a single fuzzy probability distribution for a component from several possible distributions by considering particular features that map to the Weibull parameters. The use of fuzzified probabilities in fault trees has been considered by others; once the fuzzy component failure probabilities have been calculated, a fuzzy system reliability distribution can be obtained which portrays the inexact nature of the data on which the result is based.

Using the coefficient of correlation in method-comparison studies.

The coefficient of correlation (R) is one of the most commonly computed statistics in method-comparison studies. Usually, it is simply quoted without interpretation. In this paper, we show how R may be used to detect interference, nonlinearity, and misuse of the imprecision components. Specifically, one may precisely predict what R should be by considering the imprecisions of the two methods being compared, even before the comparison is performed. When the actual R disagrees with the predicted R, then one of the mentioned effects is present. We also describe a statistical test to detect these effects at the P = 0.05 level, then evaluate this test by using computer simulation and present two examples of its use. We also present the theory underlying the usage of R, including how R is affected by the distribution and range of the data, by the joint imprecisions of the methods being compared, by the sample size, and by the randomness of the specimen-selection process.