• Analysis of security and privacy risks in large language models 

  ȘARAPOVA, Ana (Universitatea Tehnică a Moldovei, 2025)

  The thesis consists of an introduction, three chapters, a conclusion, a list of references, and a list of appendices. The research begins with an introduction into the domain of artificial intelligence (AI) systems followed ...
• Development of intelligent generative models for melodies. Melody generation 

  TCACENCO, Igor (Universitatea Tehnică a Moldovei, 2025)

  The purpose of this work is to explore the integration of deep learning models, particularly LSTMs, in generating harmonically coherent and stylistically diverse chord progressions based on established music theory. To ...
• Recognition and visualization of data of an automated surveillance system 

  PUȘCAȘ, Dumitru (Universitatea Tehnică a Moldovei, 2025)

  This paper presents the application of the YOLO v8 (You Only Look Once) model for real-time object detection within an automated video surveillance system. YOLO v8, recognized for its superior balance between speed and ...
• Research on methods for detecting fake news using Natural Language Processing and machine learning 

  PICIRIGA, Bogdan (Universitatea Tehnică a Moldovei, 2025)

  The spread of fake news has become a significant global challenge, particularly in the age of digital communication and social media. This research explores the application of Artificial Intelligence and Natural Language ...
• Application of computer vision in modern automated video surveillance systems 

  PETRANIS, Ana (Universitatea Tehnică a Moldovei, 2025)

  This master's thesis focuses on the development and analysis of an automated video surveillance system powered by computer vision, aimed at enhancing security and monitoring capabilities. The project integrates state-of-the-art ...
• Solutions to improve cognitive assistance for people with mental disabilities based on neural networks 

  CERLAT, Pavel (Universitatea Tehnică a Moldovei, 2025)

  The research involves the use of a neural network doodle and gesture recognition with emotional state detection to assist mentally disabled people. It involves the designing of the use of convolutional neural networks in ...
• Developing solution for banking platforms in managing medical insurances of Republic of Moldova 

  SCRINIC, Oleg (Universitatea Tehnică a Moldovei, 2025)

  This work studies strategies for integrating insurance services into banking platforms, answering the growing needs of modern financial organizations. The paper presents a complete analysis of the functional, technical, ...
• Improving vector search in Retrieval-Augmented Generation 

  MORARI, Gheorghe (Universitatea Tehnică a Moldovei, 2025)

  This thesis explores how semantic operators in a semantic algebra can improve vector search in Retrieval Augmented Generation (RAG) scenarios. Semantic operators manipulate the meaning of words and sentences in a formal, ...
• Development of generative song patterns. Creating dataset using metadata from music files 

  MAZUR, Vsevolod (Universitatea Tehnică a Moldovei, 2025)

  The purpose of this work is to study the metadata of audio files, both identification and music, from the perspective of their use for training analytical and generative neural networks. Creation of a program for extracting ...
• Analysis and research of barriers to cloud tehnologies implementation in small and medium-sized enterprises 

  HAMUD, Tarek (Universitatea Tehnică a Moldovei, 2025)

  Cloud computing is a major technological innovation that has revolutionized the way organizations and individual users manage, store and access data and IT resources. Essentially, cloud computing enables access to technology ...
• Analysis and implementation of stylometric features for intrinsic plagiarism detection systems 

  FURDUI, Alexandru (Universitatea Tehnică a Moldovei, 2025)

  This thesis investigates the analysis and implementation of stylometric features for intrinsic plagiarism detection systems, addressing the increasing demand for effective methods to detect unoriginal content in written ...
• Performance evaluation of machine learning models for automated ticket routing systems (DistilBERT, DistilRoBERTa, XLM-RoBERTa) 

  DUMBRĂVEANU, Ecaterina (Universitatea Tehnică a Moldovei, 2025)

  This thesis focuses on the review and performance testing of machine learning models for automated ticket routing systems, an essential instrument for raising customer satisfaction and operational effectiveness in IT support ...
• Performance evaluation of machine learning models for automated ticket routing systems (DistilBERT multilingual cased, InfoXLM, DistilBERT base uncased) 

  DEREVENCO, Serghei (Universitatea Tehnică a Moldovei, 2025)

  Automated ticket routing systems are vital for efficiently managing customer and IT support inquiries. This thesis evaluates the performance of three prominent Natural Language Processing (NLP) models: DistilBERT-base-uncased, ...
• Research on microservices architecture for an Automated Surveillance System 

  CURNICOV, Artur (Universitatea Tehnică a Moldovei, 2025)

  This work presents an exploration of designing an Automated AI Surveillance System using a microservices architecture. The study begins with an analysis of the research domain, providing an overview of automated surveillance ...
• Analysis and research of computer vision methods for object recognition in images 

  CERNEV, Evgheni (Universitatea Tehnică a Moldovei, 2025)

  This work describes a mobile application for bartenders that utilizes Computer vision techniques and neural networks to identify drink ingredients. This research seeks to develop a system that can present a selection of ...
• Research of stylometry features for intrinsic plagiarism detection systems 

  CAZACU, Constantin (Universitatea Tehnică a Moldovei, 2025)

  The report is structured into three chapters, each addressing a crucial aspect of the work process. The initial chapter serves as an introduction, presenting the identified problem that necessitates resolution. This chapter ...
• Integration of artificial intelligence techniques in the healthcare system in Moldova 

  BASSO, Nicolae (Universitatea Tehnică a Moldovei, 2025)

  Scopul acestei lucrări este de a dezvolta modele predictive bazate pe inteligența artificială pentru identificarea riscului de diabet, sprijinind astfel diagnosticul timpuriu si optimizarea resurselor medicale. Lucrarea ...
• The integration of large language models into intelligent agents for enhanced interactive storytelling 

  VRABIE, Teodor (Universitatea Tehnică a Moldovei, 2024)

  Această teză explorează crearea și implementarea unui sistem interactiv de povestire prin integrarea perfectă a Unity, API-ului OpenAI și Unity Machine Learning Agents (ML-Agents). Obiectivul principal este furnizarea unei ...
• Analiza și implementarea metodelor și tehnicilor de proiectare a unui analizor de cod static 

  NANI, Victor (Universitatea Tehnică a Moldovei, 2024)

  Uneltele folosite pentru analiză statică de cod sunt esentiale si, în acelasi timp, prezintă multe elemente care încă necesită să fie explorate în lumea modernă a ingineriei soft-ware. Acestea sunt folosite din ce în ce ...
• Clasificarea automatizată a tichetelor folosind algoritmi de procesare a limbajului natural și metode de învățare profundă 

  ȚURCAN, Cătălin (Universitatea Tehnică a Moldovei, 2024)

  In short, this paper discusses the real world advantages of implementing an automated ticket classification. Before reaching the final solution, the required data must be transformed from its original state, in one better ...

View more