OJS Tashkent State University of Economics
Not a member yet
554 research outputs found
Sort by
AUDIO SPECTROGRAM TRANSFORMER (AST): ADVANTAGES OVER TRADITIONAL ALGORITHMS IN SPEECH-TO-TEXT (STT)
No full textAutomatic Speech Recognition (ASR) has seen significant advancements in recent years, largely due to the development of deep learning models. One of the most notable advancements is the Spectrogram Transformer, a variant of the Transformer architecture tailored for audio processing tasks. In this paper, we review the Spectrogram Transformer and compare it with other traditional ASR algorithms. We discuss its benefits, such as improved performance on noisy audio and better modeling of long-range dependencies. Additionally, we explore its applications in various domains, including voice assistants, transcription services, and audio indexing. Through experiments on benchmark datasets like LibriSpeech and the Speech Commands Dataset, we demonstrate the effectiveness of the Spectrogram Transformer in achieving state-of-the-art performance. Our findings suggest that the Spectrogram Transformer offers a promising direction for future advancements in ASR technology
KOMPYUTER TARMOQLARINI DDOS HUJUMLARIDAN HIMOYA QILUVCHI DASTURIY VOSITALARNING QIYOSIY TAHLILI
No full textUshbu maqolada DDOS hujumlaridan himoyalanishda keng foydalaniladigan hamda axborot xavfsizligiga javob bera oladigan vositalari keltirilgan. Shu bilan birga, ushbu dasturiy vositalarning bir-biridan afzalliklari hamda qo‘shimcha imkoniyatlari to‘g‘risida ma’lumotlar keltirilgan
STUDYING THE FACTORS AFFECTING THE CREATIVE ABILITIES OF CHILDREN USING THE ANALYSIS OF THE CORRELATION COEFFICIENT
No full textCreative abilities play a pivotal role in a child\u27s cognitive development, influencing their problem-solving skills, adaptability, and future success. This study aimed to explore the factors that affect creative abilities in children through the analysis of correlation coefficients. The correlation coefficient of the influencing factors for the creative ability to master activities, knowledge, understanding Creative abilities play a pivotal role in a child\u27s cognitive development, influencing their problem-solving skills, adaptability, and future success. This study aimed to explore the factors that affect creative abilities in children through the analysis of correlation coefficients. The correlation coefficient of the influencing factors for the creative ability to master activities, knowledge, understanding at the stage of development of the child\u27s knowledge was studied, evaluation method and algorithm were developed. parameters selected as factors influencing children\u27s creative ability the power to monotonically influence the outcome was measured. parameters are processed according to the Likert scale, the formulas of the correlation coefficient were used to calculate the influence of the matrix on the matrix. The correlation coefficient analysis revealed substantial influences of age, interest in activities, teacher\u27s skills, and family environment on children\u27s creative abilities. These factors demonstrated strong associations, emphasizing their significant impact on nurturing creative potential in children.at the stage of development of the child\u27s knowledge was studied, evaluation method and algorithm were developed. parameters selected as factors influencing children\u27s creative ability the power to monotonically influence the outcome was measured. parameters are processed according to the Likert scale, the formulas of the correlation coefficient were used to calculate the influence of the matrix on the matrix. The correlation coefficient analysis revealed substantial influences of age, interest in activities, teacher\u27s skills, and family environment on children\u27s creative abilities. These factors demonstrated strong associations, emphasizing their significant impact on nurturing creative potential in children
МОДЕЛИ ОБСЛУЖИВАНИЯ ОЧЕРЕДЕЙ НА УЗЛАХ СЕТИ ТЕЛЕКОММУНИКАЦИИ
No full textРассмотрены существующие статистические алгоритмы и потоковые модели обслуживания очередей на узлах сети телекоммуникации, приведены основные требования к методам обслуживания очередей, предложена модель динамического гарантированного обслуживания очередей, проведен сравнительный анализ вероятностно-временных характеристик предложенной и известной (приоритетной) модели обслуживания очередей
ОБЪЕКТЛАРНИНГ МУҲИМЛИЛИК ДАРАЖАЛАРИГА МОС КЛАСТЕРЛАШ ВА СИНФЛАШТИРИШ АЛГОРИТМЛАРИ
No full textМақолада белгилар қиймати аниқ сонлардан иборат бўлган объектларни кластерлаш ва синфлаштириш масаласи муфассал тадқиқ этилган. Ушбу тадқиқотда берилган барча объектлар ихтиёрий равишда r-та синфга ажратилади. Бу синфлар мажмуаси ўқув танланма деб аталади. Ўқув танланманинг йўл элементлари объектлар бўлиб, устун элементлари эса белгилар сифатида қаралади. Бизнинг ҳолатимизда белгилар аниқ сонлар орқали берилади.
Ўқув танланмани дастлабки ишлов бериш жараёнида объект белгиларини нормаллаштириш, аномал маълумотларни олиб ташлаш ва улар ўрнида нормал маълумотларни қўйиш, ҳамда ўқув танланмадан тушиб қолган маълумотларни тиклаш амалга оширилади. Мақолада ҳар бир объект тадқиқ этилади ва танланмадаги объектнинг муҳимлиги мезон орқали баҳоланади. Бу мезон объектнинг тадқиқ этилаётган синф шаклланишига қўшган ҳиссасини баҳолашга ёрдам беради. Биринчи навбатда объектнинг ўз синфига қўшган ҳиссаси, сўнгра қолган синфларни шаклланишига қўшган ҳиссаси баҳоланади. Агар тадқиқ этилаётган объектнинг қандайдир синф шаклланишига қўшган ҳиссаси юқори бўлса, ушбу объект ўша синфга ўтказилади. Бу жараён барча синф объектлари учун кетма-кет тўлиқ равишда бир неча марта амалга оширилади. Жараён объектлар жойи ўзгармасдан ва ўхшашлик даражаси керакли фоиздан ошганда тўхтатилади
ҚОРАМОЛ ЮЗ ТАСВИРЛАРИГА ДАСТЛАБКИ ИШЛОВ БЕРИШ УСУЛЛАРИ
No full textҚорамолларни идентификация қилишда юз тасвирларига дастлабки ишлов бериш жараёнлари муҳим аҳамиятга эга. Ушбу тадқиқотда қорамол юз тасвирларини тайёрлаш учун қўлланиладиган асосий алгоритмлар, уларнинг қўлланилиши ва самарадорлиги таҳлил қилинган. Таклиф этилган усуллар орасида тасвирларнинг ўлчамини нормаллаштириш, шовқинларни фильтрлаш, ранг ва контрастни яхшилаш, шунингдек, қизиқиш майдонларини нормаллаштириш каби жараёнлар мавжуд. Тадқиқот натижалари тасвирларга дастлабки ишлов бериш чуқур ўқитиш моделларининг аниқлик даражасини сезиларли даражада оширишини кўрсатади. Масалан, қайта ишланган тасвирлардан фойдаланган ҳолда, моделнинг идентификация самарадорлиги 10-15% гача яхшиланган. Шунингдек, мақолада келтирилган алгоритмлар ҳар хил сифатли маълумот тўпламларини шакллантириш учун қўлланиши мумкин бўлган универсал ечим сифатида тавсия этилади. Ушбу тадқиқот натижалари қорамолларни идентификация қилишдаги аниқликни оширишга ҳисса қўшади ва тасвирларга дастлабки ишлов бериш соҳасидаги илмий изланишлар учун янги имкониятларни очиб беради
THREAT ANALYSIS AND PROTECTIVE MEASURES FOR VOICE-BASED IDENTIFICATION
No full textThis article presents an analysis of the threats and risks that may arise in the implementation of voice-based applications. In particular, the issues of classifying threats according to the STRIDE (Spoofing, Tampering, Repudiation, Information disclosure, Denial of Service, Elevation of privileges) methodology, assessing threat risks according to the DREAD (Damage Potential, Reproducibility, Exploitability, Affected Users, Discoverability) model, and taking protective measures against them are considered [1,2]
SUN‘IY INTELLEKT ETIKASI: IJTIMOIY SOHADAGI MUAMMOLAR VA TASHABBUSLAR
No full textSun\u27iy intellektga (AI) asoslangan texnologiyalar sezilarli natijalarga erishdi, ular orasida yuzni aniqlash, o\u27z - o\u27zini boshqaradigan avtomobillar, sug\u27urta va birja aktivlarini boshqarish, mulk, xodimlarni qidirish va tanlash. Sun\u27iy intellekt iqtisodiy o\u27sish, ijtimoiy rivojlanish, odamlarning farovonligi va xavfsizligini oshirish uchun katta foyda keltiradi. Albatta, sun\u27iy intellekt va robototexnika bugungi kunda butun dunyo bo\u27ylab eng ko\u27p muhokama qilinadigan masalalar va texnologiya tendentsiyalaridan biridir
ASSESSING MACHINE LEARNING ALGORITHMS FOR CHRONIC DISEASE PREDICTION THROUGH PERFORMANCE METRICS
No full textThis research examines the efficacy of machine learning (ML) algorithms in the early detection and prediction of chronic diseases by utilizing a mix of structured and unstructured data. Chronic conditions such as diabetes and heart disease require advanced diagnostic methods due to their complex nature. Using algorithms like Support Vector Machines, Decision Trees, and Logistic Regression, this study aims to create predictive models that surpass traditional diagnostic methods. These models are rigorously tested with real-world data to ensure they are both accurate and practical for clinical use. The goal is to enhance early detection and management of chronic diseases, potentially reducing healthcare costs and improving patient outcomes. This innovative approach advances the application of artificial intelligence in healthcare, setting new standards for predictive diagnostics in chronic diseases
INTEGRATION OF AI AND BUSINESS FOR A SUSTAINABLE FUTURE
No full textThe integration of artificial intelligence (AI) into business operations presents transformative opportunities for promoting sustainability and efficiency. This paper investigates how AI technologies can be harnessed to advance sustainable business practices. Through a detailed literature review, methodological analysis, and discussion of results, this study aims to offer a robust framework for leveraging AI in pursuit of long-term sustainability goals