AN ALGORITHM FOR EXTRACTING INFORMATICS FEATURES FROM DATA USING ARTIFICIAL INTELLIGENCE METHODS.

Аннотация

Аннотация. В статье рассматриваются алгоритмы извлечения признаков информатики из данных с использованием методов искусственного интеллекта. Извлечение признаков информатики — это процесс выявления важной и полезной информации из набора данных. Этот процесс осуществляется, в частности, с использованием технологий обработки естественного языка (NLP), машинного обучения (ML) и глубокого обучения (DL). В статье подробно рассматриваются этапы сбора данных, предварительной обработки, извлечения и отбора признаков, а также алгоритмы извлечения признаков информатики с использованием методов машинного обучения и глубокого обучения (например, SVM, KNN, CNN, RNN). Выпуск журнала №-26 Часть–6_ Май –2025 81 MODERN EDUCATION AND DEVELOPMENT Ключевые слова: Искусственный интеллект, данные, информатика, алгоритм, обработка естественного языка (NLP), машинное обучение (ML), глубокое обучение (DL), технология, машинное обучение, методы глубокого обучения, SVM, KNN, CNN, RNN, текст, изображение и аудио, формат. Algorithms are used to identify important features from data in various formats, such as text, images, and audio, and are of great importance for effective data analysis and decision-making. The article provides a detailed understanding of the practical application of informatics features extraction technologies using artificial intelligence and their role in data analysis. In data analysis, extracting important information from data sets or texts is one of the main tasks of artificial intelligence (AI) technologies. Extracting information from data sets involves identifying important and useful information. This process is carried out, in particular, using natural language processing (NLP), machine learning (ML), and deep learning methods. Basic Steps in Extracting Information from Data. The basic steps in extracting information from data are as follows. Data Collection and Aggregation. The first step in extracting information from data is to collect data. Data can be in different formats: text, images, video, or audio. Specific algorithms and AI methods are used for each format. For example, NLP is used for text data, convolutional neural networks (CNN) for images, and recurrent neural networks (RNN) for audio data. Data Preprocessing. Before data extraction, the data must be cleaned, normalized, and encoded. This process includes the following. Text cleaning. Cleaning texts from stop words, special characters, and non- standard words. Lexical normalization: Lemmatization or stemming of words. Tokenization. Dividing text into small pieces (tokens). Feature Selection and Extraction. Выпуск журнала №-26 Часть–6_ Май –2025 82 MODERN EDUCATION AND DEVELOPMENT In feature selection, the main and computationally important parts of the data are separated using a specific algorithm. TF-IDF (Term Frequency-Inverse Document Frequency): Used to identify important words for textual data. This method takes into account how often a word occurs in a document and how widespread it is in the entire data set. Word Embeddings (Word2Vec, GloVe): Represents words using vectors, taking into account the semantic meaning of words, which provides understanding of words in relation to each other. Feature Selection. When selecting features, not all features are equally important. Some features may be more important than others in identifying information. The following methods are used in feature selection. Mutual Information: Measures the relationship between features. PCA (Principal Component Analysis): Combines similar features and reduces the dimensionality of the data. Algorithms and Methods. Machine Learning (ML) Algorithms. Machine learning algorithms are used to extract informational features from data. The following methods are widely used. KNN (K-Nearest Neighbors): Used to group objects with similar features. SVM (Support Vector Machine): Effective in separating categories and selecting features. Random Forest. Determines the relationship between features through many decision trees. Deep Learning Methods. Deep learning can extract informational features from complex data. Convolutional neural networks (CNN) and recurrent neural networks (RNN) are especially effective in analyzing text, image, and audio data. CNN (Convolutional Neural Networks). Used to extract features from images and video data. In each layer, important elements of the image are extracted. RNN (Recurrent Neural Networks). Used to extract features from sequential data, such as text or time series data. Advanced models such as LSTM (Long Short- Выпуск журнала №-26 Часть–6_ Май –2025 83 MODERN EDUCATION AND DEVELOPMENT Term Memory) or GRU (Gated Recurrent Unit) help extract informatics features by remembering past states. Emerging Feature Analysis. The extracted features are then analyzed and based on them, model outputs or decisions are made. Machine learning or deep learning models are often used in feature analysis. Classification, regression, and other forecasting