Understanding and predicting COVID-19 clinical trial completion vs. cessation

As of March 30 2021, over 5,193 COVID-19 clinical trials have been registered through Clinicaltrial.gov. Among them, 191 trials were terminated, suspended, or withdrawn (indicating the cessation of the study). On the other hand, 909 trials have been completed (indicating the completion of the study). In this study, we propose to study underlying factors of COVID-19 trial completion vs. cessation, and design predictive models to accurately predict whether a COVID-19 trial may complete or cease in the future. We collect 4,441 COVID-19 trials from ClinicalTrial.gov to build a testbed, and design four types of features to characterize clinical trial administration, eligibility, study information, criteria, drug types, study keywords, as well as embedding features commonly used in the state-of-the-art machine learning. Our study shows that drug features and study keywords are most informative features, but all four types of features are essential for accurate trial prediction. By using predictive models, our approach achieves more than 0.87 AUC (Area Under the Curve) score and 0.81 balanced accuracy to correctly predict COVID-19 clinical trial completion vs. cessation. Our research shows that computational methods can deliver effective features to understand difference between completed vs. ceased COVID-19 trials. In addition, such models can also predict COVID-19 trial status with satisfactory accuracy, and help stakeholders better plan trials and minimize costs.

Method of adjusting fuel equipment of a diesel locomotive by heat release feature under operating conditions

Diesel engines supply mechanical power to almost half of the locomotives of the Russian railways. To ensure the passport characteristics of a diesel locomotive during the entire period of operation, periodic adjustment of the fuel supply equipment is required. When adjusting it, it is necessary to ensure the balance of power between the diesel cylinders while not exceeding the protective parameters for the maximum combustion pressure and the temperature of the exhaust gases. This is achieved due to the identity of the cyclic fuel supply through the diesel cylinders and the close correspondence of the fuel supply advance angles between different cylinders. Existing methods of tuning the fuel supply equipment don’t allow performing the adjustment with the required accuracy or are too complicated and laborious to implement. This article proposes theoretically substantiated and experimentally tested method for adjusting the cyclic supply and the fuel supply advance angle based on the results of measuring the pressure in the cylinder through a standard indicator channel under operating conditions. The indicator diagram is used to calculate the characteristics of active heat release, which are used to determine the relative parameters featuring the state of the fuel supply equipment. The performed computational study showed that the proposed parameters are equivalent to the relative cyclic fuel supply and the relative advance angle of the fuel supply. An experimental check, carried out on a single-cylinder compartment of OCH18/22 engine with hydromechanical fuel equipment, showed the possibility of adjusting the fuel supply equipment by the proposed method with satisfactory accuracy.

Structural Crack Detection and Recognition Based on Deep Learning

The detection and recognition of surface cracks are of great significance for structural safety. This paper is based on a deep-learning methodology to detect and recognize structural cracks. First, a training dataset of the model is built. Then, three neural networks, AlexNet, VGGNet13, and ResNet18, are employed to recognize and classify crack images. The tests indicate that the ResNet18 model generates the most satisfactory results. It is also found that the trained YOLOv3 model detects the crack area with satisfactory accuracy. This study also confirms that the proposed deep learning as a novel technology has the potential to be an efficient and robust tool for crack detection and recognition to replace traditional methods.

Application of a terminal-ballistics model for estimating munition lethal radius on mortar projectiles and rocket warheads

In our previous work, we developed a new terminal-ballistics model developed for artillery high explosive projectiles with natural fragmentation. Lethal radius is here defined as the distance from the detonation point where the effective fragment (80 J) density is 1 frag/m2. In the model, the vertical position of the projectile upon impact is assumed, which means that the lethal radius defines a circular lethal zone (maximum lethal zone of the projectile on the ground). In the research, presented in this paper, we have applied the model to mortar projectiles and rocket warheads, which have different designs compared with classic artillery projectiles. The results from the model, compared to the available experimental results from the quarter-circular arena (used in our country), showed satisfactory accuracy for these types of munitions.

To the issue of monitoring the length of degasification boreholes

Based on the application of the echolocation method, the length of degassing wells at the coal mine of the Leninsky geological and economic district of Kuzbass, drilled into the coal seam from mining workings, was measured. Recommendations for the effective use of geophysical equipment for monitoring the state of degassing wells in coal mines are justified and formed. The results of measurements of the length of horizontal degassing wells drilled from mine workings, as well as the methane content in them, are presented. Software has been developed to refine the performed measurements of the length of degassing wells, which allows you to quickly determine the length of degassing wells with satisfactory accuracy. The results of the well length processed with the help of the software are presented.

Methods For Calculating Radio Holograms Of Volumetric Objects

A method for calculating holograms for volumetric objects based on the representation of objects in the form of ensembles of virtual point sources distributed on a set of parallel planes has been proposed. The proposed method is the development of the well-known method in which objects are represented as ensemble of real point scatterers. The possibilities of the proposed method are demonstrated by calculating a hologram of a fragment of a sphere, on which 1000 points are randomly selected, at which radiation emanating from the center of the sphere is scattered. The choice of a fragment of a sphere as an object under study is due to the fact that when calculating its hologram, phase errors inherent in approximate calculations are most pronounced. The calculations were performed for the frequency range of 2...100 GHz, the sphere radius of 0.5 m, a two-dimensional hologram size of 0.65×0.65 m, and a pixel count of 512×512. It is shown that, in comparison with the known method, the proposed method makes it possible to calculate the amplitude of a hologram with satisfactory accuracy if virtual sources are placed on parallel planes in an amount of more than 64 pieces. In the case of objects that require representation in the form of an ensemble of point scatterers in the amount of more than 1000 pieces, the calculation of their holograms by the proposed method turns out to be much more efficient than the known method.

Proposal of a simplified criterion to estimate second order global effects in reinforced concrete buildings

This work proposes a new simplified parameter for the calculation of second order global effects, based on the Galerkin’s Method by Weighted Residuals. The proposed criterion was analysed based on 21 planar frames associated with shear wall, reaching results that present satisfactory accuracy compared to the second order global analysis, even for cases where the γz coefficient is greater than 1.30.

Design of Obstacle Detection System for Visually Challenged People

This paper describes a obstacle detection system for visually impaired people using Image processing in MATLAB.This system, together with ultra-sonic sensor interfaced with Arduino detects stairs and doors with or without signage and distance of these objects from the user. This information is conveyed to the user through a speaker. The results show satisfactory accuracy in detecting stairs and extracting different signage on doors such as that of washroom, exit, elevator etc.

Experimental Studies of the Developed Method for Determining the Insulation Parameters in an Asymmetric Network by a Voltage up to 1000 V at Mining Enterprises

In the work experimental studies of the developed method for determining the insulation parameters in an asymmetric electrical network by a voltage up to 1000 V at mining enterprises were carried out. It is based on measuring the module values of line voltage, angle of phase shift between vectors of damaged phase voltage relative to the ground and line voltage at two other phases, as well on measuring the module value of damaged phase voltage relative to the ground before and after connecting additional capacitive conductivity between the network’s phases and the ground. The results of experiment showed that the developed method for determining the insulation parameters has satisfactory accuracy and safety when operating three-phase electrical networks by voltage up to 1000 V at mining enterprises.

Addressing Complete New Item Cold-Start Recommendation: A Niche Item-Based Collaborative Filtering via Interrelationship Mining

Recommender system (RS) can be used to provide personalized recommendations based on the different tastes of users. Item-based collaborative filtering (IBCF) has been successfully applied to modern RSs because of its excellent performance, but it is susceptible to the new item cold-start problem, especially when a new item has no rating records (complete new item cold-start). Motivated by this, we propose a niche approach which applies interrelationship mining into IBCF in this paper. The proposed approach utilizes interrelationship mining to extract new binary relations between each pair of item attributes, and constructs interrelated attributes to rich the available information on a new item. Further, similarity, computed using interrelated attributes, can reflect characteristics between new items and others more accurately. Some significant properties, as well as the usage of interrelated attributes, are provided in detail. Experimental results obtained suggest that the proposed approach can effectively solve the complete new item cold-start problem of IBCF and can be used to provide new item recommendations with satisfactory accuracy and diversity in modern RSs.