Ahmet Yesevi University

Yazılı kaynaklarda isimlerine rastladığımız ancak günümüzde pek çoğunun ören yeri halinde kaldığı hatta mimari eserlerinin yok olma derecesinde toprak altında bulunduğu Sir Derya havzasında sıralanan Taraz, Sayram (İsficab), Otrar, Yesi (Türkistan), Sidak, Savran, Yenikent, Cend, Kumkent, gibi şehirler Oğuz ve Kıpçak Türklerinin yerleşim yerleri arasındadır. Savaşlar, göçler, doğal erozyonlar ve Sovyet döneminde yapıların tahrip edilmesi, bu kentlerin günümüze harabe görünümünde ulaşmasına sebep olmuştur. Kalıntı halindeki bu yerleşimlerden biri de çalışmamıza konu olan Sığanak’tır. Kentte 2020’li yıllarda gerçekleştirilen kazı faaliyetleri sonucunda birçok yeni bilgi ortaya konmuştur. Bunlar içerisinde bölgede tespit edilen iki mezar yapısı hakkındaki veriler yeni yorumlarla metin içerisinde aktarılmıştır. Mezar yapılarının kazılar neticesinde türbe olduğu anlaşılmıştır. Türbelerin bölge için öneminin anlaşılabilmesi açısından araştırmanın ilk safhasında Sığanak kentinin tarihi kaynaklardaki yeriyle birlikte coğrafi konumu ve türbe mimarisinin gelişimine kısaca değinilmiştir. Ardından kentte yürütülen arkeolojik kazı çalışmalarında tespit edilen iki türbeyle alakalı veriler ortaya konulmuştur. Metin içerisinde yapıların hüviyeti, plân tipleri, bezeme özellikleri, kabir özellikleri, defnedilen kişilerle ilgili bilgilere yer verilerek, yapıların karşılaştırmaları yakın coğrafyadaki örnekler üzerinden yapılmıştır. Ayrıca bezeme özellikleri ve tarihlendirmeleriyle ilgili son veriler açıklanmıştır.

This article analyzes the scientific foundations and practical effectiveness of applying machine learning and deep learning methods for the automatic detection and severity assessment of dysarthria. Dysarthria is a common symptom of neurological disorders such as Parkinson’s disease, cerebrovascular accidents, cerebral palsy, and multiple sclerosis, leading to impairments in articulatory, prosodic, and resonant components of speech. Due to the subjectivity and time consumption of traditional clinical diagnostics, the demand for automated speech processing systems is increasing. In this study, the UA-Speech and TORGO datasets were employed, and classical machine learning models such as RandomForest, XGBoost, LightGBM, SVM, LogisticRegression, and MLPClassifier were compared with deep learning approaches based on CNN architectures. The models were optimized using GridSearchCV, while the pronounced data imbalance was addressed with ADASYN and class weight techniques. Key acoustic features such as MFCC, Mel spectrograms, and pitch were used to construct the feature space. The results indicate that machine learning models can achieve high accuracy in detecting dysarthria symptoms, while hybrid approaches combining CNNs provide improved overall performance. This research lays the groundwork for developing automated early dysarthria diagnostic systems and contributes to the advancement of technologies aimed at clinical practice.

This research aims to analyze the applicability and performance of face recognition technologies using different Python-based libraries. Face recognition and comparison tests were performed using the face_recognition, OpenCV and DeepFace libraries, respectively . First, basic face detection application was successfully implemented with the face_recognition library. Then, face recognition was performed using traditional methods using OpenCV. But limited accuracy results were obtained. Finally, deep learning-based face recognition was performed with the DeepFace library, high accuracy rates were achieved, but high processing power was required. All three methods were systematically compared in terms of hardware requirements, ease of use, processing time and accuracy rates. The findings revealed that each library has advantages and limitations for certain application scenarios. In this context, it was emphasized that the selection of the appropriate algorithm for the needs in the design of face recognition systems is critical.

The aim of the work was to study the changes in social consciousness in Kazakhstan over the past 50 years in the context of social modernization and globalization. To achieve the goal, a mixed method of analysis was used, including the development of an original questionnaire of 30 questions and in-depth interviews. The study involved 500 respondents from Kazakhstan representing three age groups: young generation (18-29 years old), middle age (30-50 years old) and older generation (51 years old and older). The results of the study showed significant differences in the perception of key social processes by different generations. Thus, 78% of young people positively assess the introduction of new technologies, seeing them as an opportunity for professional growth, improved quality of life and greater access to education. Among the older generation, only 39% of respondents expressed a positive attitude to technological changes, which is due to difficulties in adaptation and adherence to traditional ways of interaction. Gender analysis revealed consistent differences in the perception of gender roles. Thus, 65% of women from the older generation consider gender roles to be unchanged, reflecting adherence to traditional values, while 72% of young people support their transformation and advocate equal opportunities in professional and social spheres. The data collected allowed for a detailed comparative analysis showing how social attitudes have changed under the influence of technological progress, social reforms and global processes.

This study aims to simulate the tactics and techniques of APT38 (Lazarus), one of the prominent cyber threat actors, as defined within the MITRE ATT&CK framework, in a virtualized corporate network environment, and to analyze the detection capabilities of existing security solutions against these attacks. During the simulation, the attacker gains access to the system by exploiting a vulnerability on the web server (web shell), then performs privilege escalation by leveraging a PrintSpooler vulnerability, and finally achieves lateral movement within the network using the compromised account to access sensitive data. The virtual infrastructure consists of DMZ and LAN segments and includes security solutions such as PfSense, Suricata, Splunk, and Carbon Black. Each step in the scenario is documented to correspond with relevant MITRE ATT&CK techniques. This study demonstrates the extent to which APT38's attack methods can be detected by security systems and presents improvements for activities that are not detected.