глибинне навчання

У статті проведено дослідження новітнього напрямку у штучному інтелекті - Великі Мовні Моделі, які відкривають нову еру в обробці природної мови, надаючи можливість створення більш гнучких і адаптивних систем. З їх допомогою досягається високий рівень розуміння контексту, що збагачує досвід користувачів та розширює сфери застосування штучного інтелекту. Великі мовні моделі мають величезний потенціал для переосмислення взаємодії людини з технологіями та зміни уявлення про машинне навчання.

In agriculture, early detection of crop diseases is imperative for sustainability and maximizing yields.  Rooted in Agriculture 4.0, our innovative approach  combines pre-trained Convolutional Neural Networks (CNNs) models with data-driven solutions to address global challenges related to water scarcity.  By integrating the combined $L_{1}/L_{2}$ regularization technique to our model layers, we enhance their flexibility, reducing the risk of the overfitting effect of the model.  In the orange dataset used in our experiments, we have 1790 orange images, including a class

The COVID-19 showed us that rapid and accurate diagnostics is a necessity.  Therefore, researchers began to implement deep learning models that can help the doctors to reach faster and reliable results, but there are more development to be done.  In our research paper, we introduced an innovative approach to enhance the Vision Graph model's accuracy for better results.  Our method exploits the strength of the ConvMixer architecture and Attention mechanism.  We start by utilizing Depthwise convolution and Pointwise convolution to capture spatial information in detail whi

The field of time series forecasting has grown significantly over the past several years and is now highly active.  In numerous application domains, deep neural networks are exact and powerful.  They are among the most popular machine learning techniques for resolving big data issues because of these factors.  Historically, there have been numerous methods for accurately predicting the subsequent change in time series data.  The time series forecasting problem and its mathematical underpinnings are first articulated in this study.  Following that, a description of the m

The objective of this paper is to focus on the study of a periodic temporal parabolic equation involving a variable exponent $p(x)$.  After proving the existence and uniqueness of the solution, we provide a method for its numerical simulation using emerging deep learning technologies.

Among the difficult problems in mathematics is the problem of solving partial differential equations (PDEs).  To date, there is no technique or method capable of solving all PDEs despite the large number of effective methods proposed.  One finds in the literature, numerical methods such as the methods of finite differences, finite elements, finite volumes and their variants, semi-analytical methods such as the Variational Iterative Method, New Iterative Method and others.  In recent years, we have witnessed the introduction of neural networks in solving PDEs.  In this w

In this work, we are interesting in solving the 1D and 2D nonlinear stiff reaction-diffusion Brusselator system using a machine learning technique called Physics-Informed Neural Networks (PINNs).  PINN has been successful in a variety of science and engineering disciplines due to its ability of encoding physical laws, given by the PDE, into the neural network loss function in a way where the network must not only conform to the measurements, initial and boundary conditions, but also satisfy the governing equations.  The utilization of PINN for Brusselator system is stil