Defect prediction is one of the key challenges in software development and programming language research for improving software quality and reliability. The problem in this area is to properly identify the defective source code with high accuracy. Developing a fault prediction model is a challenging problem, and many approaches have been proposed throughout history. The recent breakthrough in machine learning technologies, especially the development of deep learning techniques, has led to many problems being solved by these methods. Our survey focuses on the deep learning techniques for defect prediction. We analyse the recent works on the topic, study the methods for automatic learning of the semantic and structural features from the code, discuss the open problems and present the recent trends in the field.
22 July 2014 in Izvestiya, Physics of the Solid Earth
We describe the parallel algorithms for studying the structural features of the anomalies in the gravity and magnetic fields of the lithosphere, which are based on the height transformations of the data. The algorithms are numerically implemented on the Uran supercomputer. The suggested computer technology is used for constructing the maps of the regional and local anomalies of the magnetic and gravity fields for the northeastern sector of Europe within an area confined between 48°–62° E and 60°–68° N.
One of the most significant challenges of the mining industry is resource yield estimation from visual data. An example would be identification of the rock chunk distribution parameters in an open pit. Solution of this task allows one to estimate blasting quality and other parameters of open-pit mining. This task is of the utmost importance, as it is critical to achieving optimal operational efficiency, reducing costs and maximizing profits in the mining industry. The mentioned task is known as rock fragmentation estimation and is typically tackled using computer vision techniques like instance segmentation or semantic segmentation. These problems are often solved using deep learning convolutional neural networks. One of the key requirements for an industrial application is often the need for real-time operation. Fast computation and accurate results are required for practical tasks. Thus, the efficient utilization of computing power to process high-resolution images and large datasets is essential. Our survey is focused on the recent advancements in rock fragmentation, blast quality estimation, particle size distribution estimation and other related tasks. We consider most of the recent results in this field applied to open-pit, conveyor belts and other types of work conditions. Most of the reviewed papers cover the period of 2018-2023. However, the most significant of the older publications are also considered. A review of publications reveals their specificity, promising trends and best practices in this field. To place the rock fragmentation problems in a broader context and propose future research topics, we also discuss state-of-the-art achievements in real-time computer vision and parallel implementations of neural networks.
1 July 2006 in Physics of Metals and Metallography
An algorithm for modeling the evolution of a monodispersed ensemble of precipitates of a variable composition during heat treatment has been developed. In this algorithm, we for the first time have...
The work is devoted to developing the parallel algorithms for solving the initial boundary problem for the time-fractional diffusion equation. After applying the finite-difference scheme to approximate the basis equation, the problem is reduced to solving a system of linear algebraic equations for each subsequent time level. The developed parallel algorithms are based on the Thomas algorithm, parallel sweep algorithm, and accelerated over-relaxation method for solving this system. Stability of the approximation scheme is established. The parallel implementations are developed for the multicore CPU using the OpenMP technology. The numerical experiments are performed to compare these methods and to study the performance of parallel implementations. The parallel sweep method shows the lowest computing time.
1 November 2012 in Journal of Computational Science