A persistent problem in database education is that conventional learning management systems and simple SQL autograders usually verify whether an answer is correct but do not diagnose the underlying subtopic-level misconception or provide transparent remediation guidance. This paper presents an intelligent SQL learning support system that integrates an educational ER schema, validator-based error typing, L1–L6 markup, explainable CART diagnostics, and a pilot multidimensional item-response modeling layer. The anonymized pilot dataset contained 60 students, 30 SQL/database tasks, and 2,105 labeled attempts. Expert annotation showed high agreement (Cohen's kappa = 0.833). Learning outcomes improved in both groups, but compared with the control group, the experimental group using adaptive AI feedback improved more strongly: +17.80 percentage points versus +3.79 percentage points. The gain difference was 14.01 percentage points (Welch t(47.30) = 13.86, p < 0.001, 95% CI [11.98; 16.05], Hedges g = 3.53). For error classification, CART achieved an accuracy of 0.779, macro-F1 = 0.689, weighted-F1 = 0.793, and log-loss = 0.618 on the held-out test set (n = 317). Random forest and gradient boosting produced higher predictive scores, but CART was retained as the primary diagnostic model because of its interpretability. The MIRT layer was calibrated on a 60 x 30 item-response matrix and a six-dimensional Q-matrix. The results support the feasibility of explainable SQL diagnostics while indicating that broader cross-institutional validation is required before generalization.
Білім беру технологиялары мен оқу процесінде цифрлық платформалар және орталар ажырамас бөлікке айналуда. Оқу материалдары мен тапсырмалар білім алушыларға цифрлық форматта дайындалып ұсынылғандықтан, ол білім алу және оқыту процесіне елеулі өзгерістер енгізуде. Бұл өзгерістер тек цифрлық форматтағы оқыту әдістеріне ғана емес, сонымен қатар білімді бағалауға да қатысты. Осы жұмыста SQL тілін қолдауға арналған кешенді жасанды интеллект жүйесі ұсынылған. Ол үш негізгі компонентті біріктіреді: RUBERT моделіне негізделген студенттердің қателерін түрлері бойынша жіктеу модулі, XGBoost алгоритіміне негізделген ұсынымдық модуль бұл студенттердің қателер профилі мен жұмыс қарқынына сүйене отырып қайталау тақырыбын анықтайды, тапсырмаларға жауап беру теориясының (IRT) екі параметрлі моделін қолданатын бейімделгіш тестілеу жүйесі, ол білім алушының ағымдағы қабілет деңгейіне сәйкес сұрақтардың күрделілігін динамикалық түрде таңдайды. Жүйе веб-қосымша түрінде жүзеге асырылған және «Ақпараттық жүйелер» бағытындағы 120 студенттің: SQL логтары, теориялық сұрақтарға жауаптары және тест тапсырудың уақыт белгілері сияқты деректер негізінде оқу қосымшасы (2024 жылдың қыркүйегі мен 2025жылдың мамыры аралығында) қолдану барысында сыналды.
29 March 2025 in Bulletin of Shakarim University. Technical Sciences
There are no universal algorithms for solving problems of recognizing moving objects by video surveillance systems. However, for different systems and in the case of different situations, only some specific algorithm is optimal, allowing for object recognition. This article analyzes the stability of intelligent algorithms that affect the quality of speech recognition and considers an integrated approach that integrates object detection, classification of people, and recognition of gender differences. The accumulated experience in the field of pattern recognition has allowed us to achieve high results in the creation of various devices and systems in medicine, in the industrial sector, in information processing systems and video surveillance. However, computer vision technologies and optical recognition of dynamic objects continue to be an extremely difficult part of scientific research due to the variety of video cameras and devices. As well as a wide range of applications, primarily for security purposes in crowded places, disorder detection, etc. This study presents the main tasks for developing a software system using computer vision and deep learning algorithms to identify and classify people in video streams, determine their number and determine their gender.
The relevance of the research is determined by the need to simulate wave propagation in heterogeneous media based on phononic crystals. These studies are necessary for the tasks of flaw detection, the development of non-destructive testing methods, geophysics, engineering geophysics and other methods of acoustic research. During the study, it was determined that the fragmentation of the components of two-component phononic crystals into 2, 3, … N equal parts in the unit cell leads to a catastrophic rearrangement of the dispersion relations and the appearance of new elementary wave packets for acoustic waves. Using numerical-analytical methods, it is shown that in two-component phononic crystals, the polydispersity of the components can significantly affect the average characteristics of the propagation of long acoustic waves when the polydisperse crystal is close to monodisperse with a smaller unit cell size. For polydisperse crystals close to monodisperse with the same period (unit cell size), their averaged characteristics for long waves coincide.
31 March 2025 in Bulletin of Toraighyrov University. Physics & Mathematics series
В статье изучаются вопросы, связанные с количеством поступающих в высшие учебные заведения и выделенным количеством грантов. Анализируются данные о рождаемости и данные о поступивших в вузы на грантовое обучение в Казахстане для прогнозирования выделения грантов в будущем. Расчеты проводятся интервальными периодами для выявления закономерностей между демографическими показателями и образовательным спросом. Данные включают информацию о рождаемости в двух периодах: с 1998 по 2002 годы и с 2003 по 2007 годы, и количестве выданных грантов за период с 2015 по 2019 годы и с 2020 по 2024 годы соответственно. Применяются статистические и математические методы анализа для изучения связи между числом родившихся и числом поступивших в высшие учебные заведения. Особое внимание уделяется выявлению зависимости между этими показателями. Определены основные факторы и ключевые тенденции в изменении спроса на гранты. Сравнение рождаемости и количества поступающих показывает наличие значимой связи между демографическими изменениями и количеством выделенных грантов. Предварительные результаты исследования необходимы для проведения численных расчетов прогнозирования необходимого количества грантов на предстоящий период 2025-2030 годы, а также дополнительной аналитики и разработки программного обеспечения. Ключевые слова: моделирование, прогнозирование, образовательные гранты, статистические данные, математический анализ.
Machine-learning surrogates predict composite properties but report no calibrated reliability and may violate physics. Bound-constrained conformal prediction confines intervals to physical bounds but assumes them exact. This paper keeps the corridor as a hard constraint and characterizes what a decision-support system guarantees when it is wrong. If a fraction of responses lies outside the corridor, no admissible interval covers more than , so the nominal level is attainable only if . A projection-aware calibration scores out-of-corridor points as infinitely nonconforming. Its marginal coverage is at least at every calibration size ( the fallback probability); its coverage across calibrations follows an exact Beta law. In the limit it reaches the nominal level whenever attainable; calibrate-then-clip loses up to . For the system thus delivers admissible intervals at the nominal level for a small, quantified width premium, with a feasibility verdict and a Shapley explanation. Across eight datasets it recovers the nominal level on a woven laminate with a misspecified rule-of-mixtures bound (90.3% against 85.6% for naive projection) and tracks the frontier under normative bounds the data contradict. Under exact Voigt–Reuss bounds it removes all violations and narrows intervals thirteen-fold. Width, paid inside the corridor, is the price.