Estimation Of Performance Measures Research Articles

Relevance Vector Machines for Index Direction Predictions: An Application on Borsa Istanbul

This study investigates index prediction performance of Relevance Vector Machines (RVM) and frequently applied Ridge Regression and Support Vector Machines (SVM). Daily prices of BIST Banks and BIST Financials indices of Borsa Istanbul are used to obtain one-day-ahead predictions of the algorithms. According to estimated performance measures, RVM yielded mostly the best metrics in both periods of BIST Banks. While SVM obtained the best performance metrics on BIST Financials index, metrics of RVM were not far from the best. Overall, the results indicate the applicability of RVM in predicting index directions and has a potential to be a good rival of SVM.

How much can performance measures explain of the between-cow variation in enteric methane?

Enteric CH4 produced from dairy cows contributes to the greenhouse gas emission from anthropogenic sources. Recent studies have shown that the selection of lower CH4 emitting cows is possible, but this would be simpler if performance measures already recorded on farm could be used, instead of measuring gas emission from individual cows. These performance measures could be used for selection of low emitting cows. The aim of this analysis was to quantify how much of the between-cow variation in CH4 production can be explained by variation in performance measures. A data set with 3 experiments, a total of 149 lactating dairy cows with repeated measures, was used to estimate the between-cow variation (the variation between cow estimates) for performance and gas measures from GreenFeed. The cow estimates were obtained with a linear mixed model with the diet within period effect as a fixed effect and the cow within experiment as a random effect. The cow estimates for CH4 production were first regressed on the performance and gas measures individually, and then performance and CO2 production measures were grouped in 3 subsets for principal component analysis and principal component regression. The variables that explained most of the between-cow variation in CH4 production were DMI (R2 = 0.44), among the performance measures, and CO2 production (R2 = 0.61), among gas measures. Grouping the measures increased the R2 to 0.53, when only performance measures were used, and to 0.66, when CO2 production was added to the significant performance measures. We found the marginal improvement to be insufficient to justify the use of grouped measures rather than an individual measure, since the latter avoid over fitting and simplify the model. Investigation of other measures that can be explored to increase explanatory power of between-cow variation in CH4 production is briefly discussed. Finally, the use of residual CH4 as a measure for CH4 efficiency could be considered by using either DMI or CO2 production as the sole predicting variables.

Interpretability of radiomics models is improved when using feature group selection strategies for predicting molecular and clinical targets in clear-cell renal cell carcinoma: insights from the TRACERx Renal study

BackgroundThe aim of this work is to evaluate the performance of radiomics predictions for a range of molecular, genomic and clinical targets in patients with clear cell renal cell carcinoma (ccRCC) and demonstrate the impact of novel feature selection strategies and sub-segmentations on model interpretability.MethodsContrast-enhanced CT scans from the first 101 patients recruited to the TRACERx Renal Cancer study (NCT03226886) were used to derive radiomics classification models to predict 20 molecular, histopathology and clinical target variables. Manual 3D segmentation was used in conjunction with automatic sub-segmentation to generate radiomics features from the core, rim, high and low enhancing sub-regions, and the whole tumour. Comparisons were made between two classification model pipelines: a Conventional pipeline reflecting common radiomics practice, and a Proposed pipeline including two novel feature selection steps designed to improve model interpretability. For both pipelines nested cross-validation was used to estimate prediction performance and tune model hyper-parameters, and permutation testing was used to evaluate the statistical significance of the estimated performance measures. Further model robustness assessments were conducted by evaluating model variability across the cross-validation folds.ResultsClassification performance was significant (p < 0.05, H0:AUROC = 0.5) for 11 of 20 targets using either pipeline and for these targets the AUROCs were within ± 0.05 for the two pipelines, except for one target where the Proposed pipeline performance increased by > 0.1. Five of these targets (necrosis on histology, presence of renal vein invasion, overall histological stage, linear evolutionary subtype and loss of 9p21.3 somatic alteration marker) had AUROC > 0.8. Models derived using the Proposed pipeline contained fewer feature groups than the Conventional pipeline, leading to more straightforward model interpretations without loss of performance. Sub-segmentations lead to improved performance and/or improved interpretability when predicting the presence of sarcomatoid differentiation and tumour stage.ConclusionsUse of the Proposed pipeline, which includes the novel feature selection methods, leads to more interpretable models without compromising prediction performance.Trial registrationNCT03226886 (TRACERx Renal)

DEPLOYR: a technical framework for deploying custom real-time machine learning models into the electronic medical record.

Heatlhcare institutions are establishing frameworks to govern and promote the implementation of accurate, actionable, and reliable machine learning models that integrate with clinical workflow. Such governance frameworks require an accompanying technical framework to deploy models in a resource efficient, safe and high-quality manner. Here we present DEPLOYR, a technical framework for enabling real-time deployment and monitoring of researcher-created models into a widely used electronic medical record system. We discuss core functionality and design decisions, including mechanisms to trigger inference based on actions within electronic medical record software, modules that collect real-time data to make inferences, mechanisms that close-the-loop by displaying inferences back to end-users within their workflow, monitoring modules that track performance of deployed models over time, silent deployment capabilities, and mechanisms to prospectively evaluate a deployed model's impact. We demonstrate the use of DEPLOYR by silently deploying and prospectively evaluating 12 machine learning models trained using electronic medical record data that predict laboratory diagnostic results, triggered by clinician button-clicks in Stanford Health Care's electronic medical record. Our study highlights the need and feasibility for such silent deployment, because prospectively measured performance varies from retrospective estimates. When possible, we recommend using prospectively estimated performance measures during silent trials to make final go decisions for model deployment. Machine learning applications in healthcare are extensively researched, but successful translations to the bedside are rare. By describing DEPLOYR, we aim to inform machine learning deployment best practices and help bridge the model implementation gap.

Kriging-Based Performance Measure Function Approximation Method for Hybrid Reliability-Based Design Optimization

Hybrid reliability-based design optimization (HRBDO) can provide an effective way to obtain the optimum design in the presence of both random and interval variables. HRBDO is typically described as a nested optimization model. It is computationally expensive when directly solving the HRBDO problem by the nested optimization method. To address this issue, this paper develops an efficient decoupled HRBDO method that aims at performance measure function approximation by the adaptive Kriging model (KPMFA). The proposed KPMFA method includes three main blocks, namely hybrid inverse reliability analysis, performance measure function approximation, and equivalent deterministic optimization. In KPMFA, the adaptation of the adaptive chaos control (ACC) algorithm for inverse reliability analysis that accommodates interval variables is developed. Moreover, an adaptive strategy with two-stage of enrichment for the Kriging model is developed to approximate performance measure functions on the region of interest. Then, the optimization can be proceeded using the Kriging model of performance measure functions. Finally, five illustrative HRBDO problems are investigated to demonstrate the accuracy and efficiency of the proposed KPMFA method.

Clinical First-Trimester Prediction Models for Gestational Diabetes Mellitus: A Systematic Review and Meta-Analysis.

Gestational diabetes mellitus (GDM) is a common pregnancy complication that negatively impacts the health of both the mother and child. Early prediction of the risk of GDM may permit prompt and effective interventions. This systematic review and meta-analysis aimed to summarize the study characteristics, methodological quality, and model performance of first-trimester prediction model studies for GDM. Five electronic databases, one clinical trial register, and gray literature were searched from the inception date to March 19, 2022. Studies developing or validating a first-trimester prediction model for GDM were included. Two reviewers independently extracted data according to an established checklist and assessed the risk of bias by the Prediction Model Risk of Bias Assessment Tool (PROBAST). We used a random-effects model to perform a quantitative meta-analysis of the predictive power of models that were externally validated at least three times. We identified 43 model development studies, six model development and external validation studies, and five external validation-only studies. Body mass index, maternal age, and fasting plasma glucose were the most commonly included predictors across all models. Multiple estimates of performance measures were available for eight of the models. Summary estimates range from 0.68 to 0.78 (I2 ranged from 0% to 97%). Most studies were assessed as having a high overall risk of bias. Only eight prediction models for GDM have been externally validated at least three times. Future research needs to focus on updating and externally validating existing models.

Performance analysis of flexible flow shop with parallel manufacturing cells

Performance evaluation for stochastic manufacturing systems is significant for system design optimization, but it is difficult to exactly analyse these systems with general arrival processes and service times. This paper studies a flexible flow shop with parallel manufacturing cells and models it as a queueing network. Parallel manufacturing cells and finite pallets lead to strong dependence among stages, and existing methods cannot be directly employed in this problem. Thus, approximate analysis is proposed to estimate performance measures of the flexible flow shop with parallel manufacturing cells based on two-moment approximations for a parallel-input synchronization station and an aggregation procedure. Numerical studies show the accuracy and the efficiency of the proposed method by comparing it with simulation results of experiments. This work can provide a basis to analyse and design more complex manufacturing systems.

Modeling sonar performance using J-divergence

This presentation demonstrates the use of J-divergence as a performance measure for detection in a sonar system. The inherent inaccuracies between system-level performance and the cell-level (PD,PF) detector operating point used in traditional analysis open the door to using approximate performance measures such as J-divergence. The properties of J-divergence making it an appealing choice are covered: summing to accrue J-divergence across multiple independent measurements (e.g., from multiple source signals, waveforms, or arrays), a data-processing inequality dictating that processing cannot improve J-divergence, and an asymptotic relationship to the traditional (PD,PF) operating point. Simple forward models of J-divergence are presented for matched filters and energy detectors when applied to the standard signal models in Gaussian noise. A “design” J-divergence, which is chosen by the desired qualitative performance level, is used in simple equations to obtain the “design” SNR require to achieve it. This provides a direct replacement for the detection threshold (DT) term in the sonar equation that is easier to apply and evaluate. Example applications of J-divergence are presented illustrating its utility in the modeling and adaptation of current systems as well as the design and analysis of new ones.

OVA-LEAK: Prognostic score for colo-rectal anastomotic leakage in patients undergoing ovarian cancer surgery

ObjectiveThe objective of the present study was to define and validate an anastomotic leak prognostic score based on previously described and reported anastomotic leak risk factors (OVA-LEAK: https://n9.cl/ova-leakscore) and to establish if the use of OVA-LEAK score is better than clinical criteria (surgeon's choice) selecting anastomosis to be protected with a diverting ileostomy. Material & methodsThis is a retrospective, multicentre cohort study that included patients who underwent cytoreductive surgery for primary advanced or relapsed ovarian cancer with colorectal resection and anastomosis between January 2011 and June 2021. Data from patients already included in the previous predictive model were not considered in the present analysis. To validate the performance of our logistic regression model, we used the OVA-LEAK formula (Annex I: https://n9.cl/ova-leakscore) for estimating leakage probabilities in a new independent cohort. Then, receiver operating characteristic (ROC) analysis was performed and area under the curve (AUC) was used to measure the performance of the model. Additionally, the Brier score was also estimated. 95% confidence intervals (CI) for each of the estimated performance measures were also calculated. Results848 out of 1159 recruited patients were finally included in the multivariable logistic regression model validation. The AUC of the new cohort was 0.63 for predicting anastomotic leak. Considering a cut-off point of 22.1% to be ‘positive’ (to get a leak) this would provide a sensitivity of 0.45, specificity of 0.80, positive predictive value of 0.09 and negative predictive value of 0.97 for anastomotic leak.If we consider this cut-off point to select patients at risk of leak for bowel diversion, up to 22.5% of the sampled patients would undergo a diverting ileostomy and 47% (18/40) of the anastomotic leaks would be ‘protected’ with the stoma. Nevertheless, if we consider only the ‘clinical criteria’ for performing or not a diverting ileostomy, only 12.5% (5/40) of the leaks would be ‘protected’ with a stoma, with a rate of diverting ileostomy of up to 24.3%. ConclusionsCompared with subjective clinical criteria, the use of a predictive model for anastomotic leak improves the selection of patients who would benefit from a diverting ileostomy without increasing the rate of stoma use.

Routing and staffing in emergency departments: A multiclass queueing model with workload dependent service times

Efficient patient flow through an emergency department is a critical factor that contributes to a hospital’s performance, which influences overall patient health outcomes. In this work, we model a multiclass multiserver queueing system where patients of varying acuity receive care from one of several wards, each ward is attended by several nurses who work as a team. Supported by empirical evidence that a patient’s time-in-ward is a function of the nurse-patient ratio in that ward, we incorporate state-dependent service times into our model. Our objective is to reduce patient time in system and to control nurse workload by jointly optimizing patient routing and nurse allocation decisions. Due to the computational challenges in formulating and solving the queueing model representation, we study a corresponding deterministic fluid model which serves as a first-order approximation of the multiclass queueing model. Next, we formulate and solve an optimization model using the first-order control equations and input the results into a discrete-event simulation to estimate performance measures, such as patient length-of-stay and ward workload. Finally, we present a case study using retrospective data from a real hospital which highlights the importance of accounting for nurse workload and service behavior in developing routing and staffing policies.

Care process optimization in a cardiovascular hospital: an integration of simulation–optimization and data mining

To provide health services, hospitals consume electrical power and contribute to the CO2 emission. This paper aims to develop a modelling approach to optimize hospital services while reducing CO2 emissions. To capture treatment processes and the production of carbon dioxide, a hybrid method of data mining and simulation–optimization techniques is proposed. Different clustering algorithms are used to categorize patients. Using quality indicators, clustering methods are evaluated to find the best cluster sets, and then patients are categorized accordingly. Discrete-event simulation is applied to each patient category to estimate performance measures such as number of patients being served, waiting times, and length of stay, as well as the amount of CO2 emission. To optimize performance measures of patient flow, metaheuristic searches have been used. The dataset of Bushehr Heart Hospital is considered as a case study. Based on K-means, K-medoid, Hierarchical clustering, and Fuzzy C-means clustering methods, patients are categorized into two groups of high-risk and low-risk patients. The number of patients being served, total waiting time, length of stay, and CO2 emitted during care processes are improved for both groups. The proposed hybrid method is an effective method for hospitals to categorize patients based on care processes. The problems and the proposed solution approach reported in this study could be applicable to other hospitals, worldwide to help both optimize the patient flow and minimize the environmental consequences of care services.

Dynamic Vehicle Allocation Policies for Shared Autonomous Electric Fleets

In the future, vehicle sharing platforms for passenger transport will be unmanned, autonomous, and electric. These platforms must decide which vehicle should pick up which type of customer based on the vehicle’s battery level and customer’s travel distance. We design dynamic vehicle allocation policies for matching appropriate vehicles to customers using a Markov decision process model. To obtain the model parameters, we first model the system as a semi-open queuing network (SOQN) with multiple synchronization stations. At these stations, customers with varied battery demands are matched with semi-shared vehicles that hold sufficient remaining battery levels. If a vehicle’s battery level drops below a threshold, it is routed probabilistically to a nearby charging station for charging. We solve the analytical model of the SOQN and obtain approximate system performance measures, which are validated using simulation. With inputs from the SOQN model, the Markov decision process minimizes both customer waiting cost and lost demand and finds a good heuristic vehicle allocation policy. The experiments show that the heuristic policy is near optimal in small-scale networks and outperforms benchmark policies in large-scale realistic scenarios. An interesting finding is that reserving idle vehicles to wait for future short-distance customer arrivals can be beneficial even when long-distance customers are waiting.

Algorithms for Queueing Systems with Reneging and Priorities Modeled as Quasi-Birth-Death Processes

We consider a Markovian multiserver queueing system with two customer classes, preemptive priorities, and reneging. We formulate this system as an infinite level-dependent quasi-birth-death process (LDQBD). We introduce an algorithm that endogenously truncates the level and calculates lower and upper bounds on stationary probabilities of this LDQBD such that the gap between the bounds can be any desired amount. Our algorithm can be applied to any LDQBD for which the rate matrices become elementwise nonincreasing above some level. This appears to be the first algorithm that provides bounds on stationary probabilities for an infinite-level LDQBD. To obtain these bounds, the algorithm first obtains lower and upper bounds on the rate matrices of the LDQBD using a novel method, which can be applied to any LDQBD. We then extend this algorithm to approximate performance measures of the system of interest and calculate exact lower and upper bounds for those that can be expressed as probabilities, such as the probability that an incoming low-priority customer will wait. We generate a wide range of instances with up to 100 servers and compare the solution times and accuracy of our algorithm with two existing algorithms. These numerical experiments indicate that our algorithm is faster than the other two algorithms for a given accuracy requirement. We investigate the impact of changing service rates on the proportion of low-priority customers served and their wait time, and we demonstrate how ignoring one of these measures can possibly mislead decision makers. Summary of Contribution: We contribute to operations research by modeling a practically important queueing system and developing an algorithm to accurately compute performance measures for that system. We also contribute to computer science by providing error and complexity analysis for the algorithm to solve a broad class of two-dimensional Markov chains with infinite state space.

Predicting relapse or recurrence of depression: systematic review of prognostic models.

Relapse and recurrence of depression are common, contributing to the overall burden of depression globally. Accurate prediction of relapse or recurrence while patients are well would allow the identification of high-risk individuals and may effectively guide the allocation of interventions to prevent relapse and recurrence. To review prognostic models developed to predict the risk of relapse, recurrence, sustained remission, or recovery in adults with remitted major depressive disorder. We searched the Cochrane Library (current issue); Ovid MEDLINE (1946 onwards); Ovid Embase (1980 onwards); Ovid PsycINFO (1806 onwards); and Web of Science (1900 onwards) up to May 2021. We included development and external validation studies of multivariable prognostic models. We assessed risk of bias of included studies using the Prediction model risk of bias assessment tool (PROBAST). We identified 12 eligible prognostic model studies (11 unique prognostic models): 8 model development-only studies, 3 model development and external validation studies and 1 external validation-only study. Multiple estimates of performance measures were not available and meta-analysis was therefore not necessary. Eleven out of the 12 included studies were assessed as being at high overall risk of bias and none examined clinical utility. Due to high risk of bias of the included studies, poor predictive performance and limited external validation of the models identified, presently available clinical prediction models for relapse and recurrence of depression are not yet sufficiently developed for deploying in clinical settings. There is a need for improved prognosis research in this clinical area and future studies should conform to best practice methodological and reporting guidelines.

Spectrum Bias and Individual Strengths of SARS-CoV-2 Serological Tests-A Population-Based Evaluation.

Antibody testing for determining the SARS-CoV-2 serostatus was rapidly introduced in early 2020 and since then has been gaining special emphasis regarding correlates of protection. With limited access to representative samples with known SARS-CoV-2 infection status during the initial period of test development and validation, spectrum bias has to be considered when moving from a “test establishment setting” to population-based settings, in which antibody testing is currently implemented. To provide insights into the presence and magnitude of spectrum bias and to estimate performance measures of antibody testing in a population-based environment, we compared SARS-CoV-2 neutralization to a battery of serological tests and latent class analyses (LCA) in a subgroup (n = 856) of the larger population based TiKoCo-19 cohort (n = 4185). Regarding spectrum bias, we could proof notable differences in test sensitivities and specificities when moving to a population-based setting, with larger effects visible in earlier registered tests. While in the population-based setting the two Roche ELECSYS anti-SARS-CoV-2 tests outperformed every other test and even LCA regarding sensitivity and specificity in dichotomous testing, they didn’t provide satisfying quantitative correlation with neutralization capacity. In contrast, our in-house anti SARS-CoV-2-Spike receptor binding domain (RBD) IgG-ELISA (enzyme-linked-immunosorbant assay) though inferior in dichotomous testing, provided satisfactory quantitative correlation and may thus represent a better correlate of protection. In summary, all tests, led by the two Roche tests, provided sufficient accuracy for dichotomous identification of neutralizing sera, with increasing spectrum bias visible in earlier registered tests, while the majority of tests, except the RBD-ELISA, didn’t provide satisfactory quantitative correlations.

Third-Party Data Fusion to Estimate Freeway Performance Measures

Transportation agencies monitor freeway performance using various measures such as VMT (vehicle-miles traveled), VHD (vehicle-hours of delay), and VHT (vehicle-hours traveled). They typically rely on data from point detectors to estimate these freeway performance measures. Point detectors such as inductive loops cannot capture the travel time for a corridor, leading to inaccurate performance measure estimation. This research develops a hybrid method, which estimates freeway performance measures using a mix of probe vehicle data provided by third-party vendors and data from traditional point detectors. Using a simulated model of a freeway (Interstate-210), the overall framework using multiple data sources is evaluated and compared with the traditional point detector-based estimation method. In the traditional method, point speeds are estimated with the flow and occupancy values using g-factors. Data from 5% of the total vehicles are used to generate the third-party provided travel time data. The analysis is conducted for multiple scenarios, including peak and off-peak periods. Results suggest that fusing probe vehicle data from third-party vendors with data from point detectors can help transportation agencies estimate performance measures better than the traditional method, in scenarios that have noticeable traffic demand on freeways.

Prognostic models for predicting relapse or recurrence of major depressive disorder in adults.

Relapse (the re-emergence of depressive symptoms after some level of improvement but preceding recovery) and recurrence (onset of a new depressive episode after recovery) are common in depression, lead to worse outcomes and quality of life for patients and exert a high economic cost on society. Outcomes can be predicted by using multivariable prognostic models, which use information about several predictors to produce an individualised risk estimate. The ability to accurately predict relapse or recurrence while patients are well (in remission) would allow the identification of high-risk individuals and may improve overall treatment outcomes for patients by enabling more efficient allocation of interventions to prevent relapse and recurrence. To summarise the predictive performance of prognostic models developed to predict the risk of relapse, recurrence, sustained remission or recovery in adults with major depressive disorder who meet criteria for remission or recovery. We searched the Cochrane Library (current issue); Ovid MEDLINE (1946 onwards); Ovid Embase (1980 onwards); Ovid PsycINFO (1806 onwards); and Web of Science (1900 onwards) up to May 2020. We also searched sources of grey literature, screened the reference lists of included studies and performed a forward citation search. There were no restrictions applied to the searches by date, language or publication status . We included development and external validation (testing model performance in data separate from the development data) studies of any multivariable prognostic models (including two or more predictors) to predict relapse, recurrence, sustained remission, or recovery in adults (aged 18 years and over) with remitted depression, in any clinical setting. We included all study designs and accepted all definitions of relapse, recurrence and other related outcomes. We did not specify a comparator prognostic model. Two review authors independently screened references; extracted data (using a template based on the CHecklist for critical Appraisal and data extraction for systematic Reviews of prediction Modelling Studies (CHARMS)); and assessed risks of bias of included studies (using the Prediction model Risk Of Bias ASsessment Tool (PROBAST)). We referred any disagreements to a third independent review author. Where we found sufficient (10 or more) external validation studies of an individual model, we planned to perform a meta-analysis of its predictive performance, specifically with respect to its calibration (how well the predicted probabilities match the observed proportions of individuals that experience the outcome) and discrimination (the ability of the model to differentiate between those with and without the outcome). Recommendations could not be qualified using the GRADE system, as guidance is not yet available for prognostic model reviews. We identified 11 eligible prognostic model studies (10 unique prognostic models). Seven were model development studies; three were model development and external validation studies; and one was an external validation-only study. Multiple estimates of performance measures were not available for any of the models and, meta-analysis was therefore not possible. Ten out of the 11 included studies were assessed as being at high overall risk of bias. Common weaknesses included insufficient sample size, inappropriate handling of missing data and lack of information about discrimination and calibration. One paper (Klein 2018) was at low overall risk of bias and presented a prognostic model including the following predictors: number of previous depressive episodes, residual depressive symptoms and severity of the last depressive episode. The external predictive performance of this model was poor (C-statistic 0.59; calibration slope 0.56; confidence intervals not reported). None of the identified studies examined the clinical utility (net benefit) of the developed model. Of the 10 prognostic models identified (across 11 studies), only four underwent external validation. Most of the studies (n = 10) were assessed as being at high overall risk of bias, and the one study that was at low risk of bias presented a model with poor predictive performance. There is a need for improved prognostic research in this clinical area, with future studies conforming to current best practice recommendations for prognostic model development/validation and reporting findings in line with the Transparent Reporting of a multivariable prediction model for Individual Prognosis Or Diagnosis (TRIPOD) statement.

Multi-objective Design and Optimization of Power Electronics Converters With Uncertainty Quantification—Part II: Model-Form Uncertainty

The second part of the multi-objective design and optimization with parametric and model-form uncertainty quantification (MDO with P&MF-UQ) is dedicated to incorporating MF-UQ into the MDO framework. MF-UQ is used to estimate the error due to modeling inaccuracies and to validate the mathematical models used in the MDO framework. In this approach, the sensitivity index introduced in the first part of the article is modified too so that it is a quantitative measure of system design robustness with regards to modeling inaccuracies as well as the manufacturing variability in the design of systems with multiple performance functions. The optimum design solution is finally realized by exploring the Pareto Front of the enhanced performance space, where the model-form error associated with each design is used to modify the estimated performance measures and the parametric sensitivity of each design point is considered to discern between cases and help identify the most parametrically robust of the Pareto-optimal design solutions. To demonstrate the benefits of incorporating uncertainty quantification analysis into the design optimization from a more practical standpoint, design of a robust high-efficiency high-power-density 1.25 kW Vienna-type rectifier is used as a case study to compare the theoretical analysis with a comprehensive experimental validation. It is shown that the final design selected using the MDO with P&MF-UQ reduces the design sensitivity and system loss by 33% and 5%, respectively, compared to the optimal design selected using the conventional MDO.

Approximate performance measures for a single station two-stage reneging queue

We study a single station two-stage reneging queue with Poisson arrivals, exponential services, and two levels of exponential reneging behaviors, extending the popular Erlang A model that assumes a constant reneging rate. We derive approximate analytical formulas representing performance measures for our two-stage queue following the Markov chain decomposition approach. Our formulas not only give accurate results spanning the heavy-traffic to the light-traffic regimes, but also provide insight into capacity decisions.

Machine Learning for End Consumers

The articles in this special section examine machine learning (ML) for end consumers. ML is a discipline that grew out of artificial intelligence (AI). At a minimum, an intelligent agent needs to perceive the environment around it, deliberate, and take the best course of actions to maximize some actual or estimated performance measures. ML was originally a trait of AI that concerned training intelligent agents to perform tasks that cannot be preprogrammed. ML has received much attention recently with advances in technologies that permeate many facets of our everyday lives, e.g., autonomous vehicles, lifelike chatbots, speech synthesis and recognition, intelligent web search, financial forecasting, personal healthcare, traffic navigation, and many other consumer applications. Key enablers that have propelled ML to the forefront of AI research include availability of vast volumes of data, algorithmic advancements that have enabled effective training of deep neural networks, and accessibility and affordability of powerful computing resources. Consequently, novel learning paradigms have been developed beyond the classical discriminative supervised, unsupervised, and semisupervised approaches. Notable novel learning paradigms include reinforcement learning, transfer learning, lifelong learning, generative adversarial learning, and more.