A dual number abstraction for static analysis of Clarke Jacobians

We present a novel abstraction for bounding the Clarke Jacobian of a Lipschitz continuous, but not necessarily differentiable function over a local input region. To do so, we leverage a novel abstract domain built upon dual numbers, adapted to soundly over-approximate all first derivatives needed to compute the Clarke Jacobian. We formally prove that our novel forward-mode dual interval evaluation produces a sound, interval domain-based over-approximation of the true Clarke Jacobian for a given input region. Due to the generality of our formalism, we can compute and analyze interval Clarke Jacobians for a broader class of functions than previous works supported – specifically, arbitrary compositions of neural networks with Lipschitz, but non-differentiable perturbations. We implement our technique in a tool called DeepJ and evaluate it on multiple deep neural networks and non-differentiable input perturbations to showcase both the generality and scalability of our analysis. Concretely, we can obtain interval Clarke Jacobians to analyze Lipschitz robustness and local optimization landscapes of both fully-connected and convolutional neural networks for rotational, contrast variation, and haze perturbations, as well as their compositions.

Large crop production losses induced by global ozone stress based on interval evaluation

Global crop yield loss due to ground-level ozone (O3) concentrations is a major challenge to food security, but a dose-response association is not easy to quantify. Here, we propose using a new metric, O3 sensitivity of crop yield (Yo), to estimate yield loss under different O3 time intervals using four observational databases. The Yo metric shows a non-linear parabola with elevated atmospheric O3 for wheat, maize, rice, soybean, and assorted vegetables. Spatial heterogeneity of yield loss varies as a function of crop type and O3 intervals. Estimates of yield loss from ozone suggest recent losses (2017-2019) may reach as high as 537 million tonnes, with a significant proportion coming with lower (30-40 ppb) exposure (325 million tonnes). Our results suggest that previous research, which only included higher (>40 ppb ozone), may have had grossly underestimated the negative effect of atmospheric O3 on crop production. Suppose these results are endemic to global crop production. In that case, additional research will be necessary to reassess ozone sensitivity and dose-responses, both spatially and temporally, to determine future air pollution impacts.

A Voltage Sag Severity Evaluation Method for the System Side which Considers the Influence of the Voltage Tolerance Curve and Sag Type

Considering the influence of user equipment voltage tolerance characteristics and sag types on the evaluation results, this paper proposes a voltage sag severity evaluation method for the system side which considers the influence of the voltage tolerance curve and sag type. As such, a quantitative evaluation of the severity of voltage sag events can be achieved. Firstly, the user’s voltage tolerance curve is used to construct the comparison reference value of the energy index, in order to realize the rapid analysis of the severity of the sag event in the normal area and the abnormal area. Secondly, aiming at the problem of insufficient descriptions of the severity difference of sag events in uncertain areas, an improved energy index evaluation model combined with user tolerance characteristics is established through an interval division and interval weight calculation, so as to divide and evaluate the severity of sag events in uncertain areas. Considering the influence of the sag type on the voltage tolerance curve and user equipment, the energy index correction factor is then constructed, and the measurement function is used for an interval evaluation to obtain the ranking result of the voltage sag severity, which is more in line with the actual situation. Finally, the rationality and effectiveness of the proposed method are verified by analyzing 24 voltage sag events at a monitoring node.

Multiple Attribute Decision-Making Model for Supplier Selection in Service-Oriented Manufacturing Paradigm

This paper proposes a multiattribute decision-making model for supplier selection under the service-oriented manufacturing, which can be used to effectively evaluate each candidate supplier. The supplier selection index system under the service-oriented manufacturing is proposed, and the interval evaluation matrix is established. In view of the mixed attribute of evaluation index, we construct a method that converts mixed attribute value to interval number. In order to avoid the subjectivity of the weight and make alternatives be provided with more discrimination, we use a combination model based on the deviation function model and the interval relative entropy ranking method to evaluate each candidate supplier. Finally, an application example is given to verify the correctness and practicability of the proposed decision-making model.

QT Interval Prolongation in Patients with Systemic Sclerosis—Are the Holter ECG Recordings a Better Option for QT Interval Evaluation?

Background and Objectives: Cardiac involvement in systemic sclerosis has important consequences on patient survival. Myocardial fibrosis and microcirculation involvement can generate arrhythmic complications, which can be associated with a higher death risk. QT interval prolongation is considered as a risk factor for ectopic ventricular events and can be evaluated using standard short ECG recordings or 24-h Holter ECG recordings. Materials and Methods: 39 patients with systemic sclerosis were submitted to a standard ECG recording at admission and 24-h Holter ECG monitoring. Results: QT interval values resulted from Holter ECG monitoring are higher than the values generated by the short-term ECG recordings. Holter ECG monitoring permits the detection of ventricular ectopy in patients with no events on standard ECG. Conclusions: In patients with systemic sclerosis, 24-h Holter ECG recordings can realize a more precise evaluation of the extent of QTc interval prolongation and ventricular ectopic events associated with myocardial involvement.

QT interval evaluation associated with the use of hydroxychloroquine with combined use of azithromycin among hospitalised children positive for coronavirus disease 2019

Introduction and aim:Hydroxychloroquine alone or in combination with azithromycin has been increasingly used for patients with coronavirus disease 2019, in both children and adults. Drugs are generally well tolerated in clinical practice; however, both can cause corrected QT prolongation. We aimed to report our experience of QT interval evaluation associated with the use of hydroxychloroquine with concurrent azithromycin among children testing positive for coronavirus disease 2019.Methods:Our single-centre; retrospective, study evaluated children with coronavirus disease 2019 disease admitted to the Pediatric Department at Sancaktepe Training and Research Hospital Istanbul, Turkey from 10 March, 2020 to 10 April, 2020. The data including demographics, clinical symptoms, co-morbid diseases, laboratory, radiological findings as well as electrocardiographs of the patients were obtained from our records. Electrocardiograms were evaluated before, one day after and at the termination of the treatment.Results:21 patients aged 9 to 18 years were evaluated. The median age was 170 months (range 112–214), 51.1% of them were girls and 48.9% were boys. Their laboratory results did not reveal any abnormalities. None of them needed intensive care. We did not detect QT prolongation during or at the termination of the treatment.Conclusion:We did not detect QT prolongation during or at the termination of the treatment in our patients due to the fact that they were not severely affected by the disease. Patients were treated in our inpatient clinic and none of them required intensive care. Laboratory results were also insignificant. Furthermore, they did not need other medications.