This research work aims to develop a model of the Uzbek language using artificial neural networks to improve natural language processing (NLP) capabilities. The article classifies various methods of creating language models and presents the stages of designing a language model using architectures such as recurrent neural networks (RNN) and long short-term memory (LSTM) networks. It also discusses the evaluation index for neural language models. Additionally, the text base features used to create the language model, the proposed neural network architecture, the hyperparameters of the neural network, and the obtained results are highlighted.
| Mualliflar | Хужаяров, Илёс, Очилов, Маннон |
|---|---|
| Jurnal | Рақамли технологияларнинг назарий ва амалий масалалари |
| Nashr sanasi | 2024-10-09 |
| Jild | 7 |
| Son | 3 |
| Betlar | 76-83 |
| Til | Rus |
| DOI | 10.62132/ijdt.v7i3.199 |
DOI: 10.62132/ijdt.v7i3.199 · Maqolaning asl sahifasi
Узбекский язык, обработка естественного языка, языковая модель, набор обучающих данных, глубокое обучение, рекуррентные нейронные сети, оценка точности модели, коэффициент недоумения, Uzbek language, natural language processing, language model, training dataset, deep learning, recurrent neural networks, model accuracy assessment, perplexity coefficient
Exploration of an unknown environment by a mobile robot during SLAM processes is a fundamental challenge. In this study, an extended iteration of the Rmap algorithm is presented by modifying its Exploration submodule…
Currently, the need to use the hydraulic system in various fields is increasing. Including excavator equipment, crane machines optimize the process of moving cargo. In the FOPID system, the power of the fluid is used in…
Methods have been developed to optimize the identification of pollen grains using statistical, dynamic, textural and specific image characteristics. Mechanisms for point and nonlinear verification of the correspondence…
This research aims to develop improved methods for diagnosing diseases in cattle using the principles of neutrosophic sets and Sugeno fuzzy inference. The study proposed new algorithms and models that effectively…
In the article, a study was conducted to comprehensively study the important factors affecting the process of migration and diffusion of harmful substances and dust particles in the atmosphere. The physical and…
This article uses two image processing methods, k-means and Threshold. For the experiment, detection, quantification and analysis of green leaf spots collected mainly from online sources were considered. The purpose of…
While non-linear S-tables are important in the security of symmetric cryptosystems, their hardware implementation also requires a large number of logic elements compared to the rest of the mappings. This makes it…
The article highlights the theoretical and practical aspects of a mathematical model for predicting the movement and level of groundwater based on the studied literature. The application of this model allows for…
This paper presents an analysis of the application of the Low-Rank Adaptation method (LoRA) for the task of monolingual text generation in Uzbek. We used the T5-base, T5-Large and uzT5 models to determine which one…
The present work is devoted to the formulation and numerical solution of coupled boundary value problems of thermoelasticity with respect to deformations for a rod, a plate, and a parallelepiped. Two equivalent coupled…
Рақамли технологияларнинг назарий ва амалий масалалари — barcha maqolalar