TSI logo 25 years
LV EN
Home > About Us > Structure and Government > Abstracts

DEGREE

PROGRAMME

FACULTY

YEAR

LANGUAGE

Enhancement Strategies for Retrieval-Augmented Generation Systems

This thesis systematically explores the enhancement of Retrieval-Augmented Generation (RAG) systems within Large Language Models, emphasizing optimization of retrieval parameters and generation accuracy. We investigate optimal configurations in RAG systems, including chunk size and overlap percentages, top-k selection, query transformations, different retrieval methods, different LLMs, namely GPT-3.5-Turbo and GPT-4, discovering that a chunk size of 500 tokens generally offers the best performance. Vector search using cosine similarity emerges as the most effective retrieval method, significantly enhancing both context precision and recall across various tasks and knowledge bases. Experimentation within the CRUD-RAG framework demonstrates its applicability in diverse tasks from content creation to knowledge refinement. Our findings indicate that enhancements in retrieval settings can markedly improve the performance of RAG systems, making them more efficient and adaptable for complex information synthesis and retrieval tasks. These results affirm the potential of systematic enhancements to improve AI-driven language models in practical applications, contributing significant insights and practical approaches to the evolving landscape of RAG system research.

Author: Sigita Lapiņa

Supervisor: Dmitry Pavlyuk

Degree: Master

Year: 2024

Work Language: English

Study programme: Computer Sciences

More...

Artificial intelligence for crew scheduling in aviation sector

This research aims to relive the factors on applying AI in crew scheduling and rostering on low-cost airlines, IndiGO Airlines was chosen as base, to minimize the time and effort of human resources and increase effectiveness. This research looks into the possibility of utilizing artificial intelligence proposed model to schedule through a decision support system to reduce mistake from human intervention. Techniques employed includes of extensive literature searches, six qualitative interviews with two respondents per industry and 109 quantitative online surveys for crew scheduling department respondents. The data collected from the survey was analyzed and presented in the form of graphs to ease interpretation of the this research by concentrating on the challenges and costs involved in Artificial Intelligence. Thus, the techniques such as Data reduction and abduction logic have been used to find the sound information out of the whole set of information.From the survey results and interview questions, there are major benefits of incorporating AI in crew scheduling and rostering. The study also presents the best approach that low-cost airlines can adopt to lower errors and uphold performance, effectively showing that the adoption of AI in the industry is significantly beneficial

Author: Slavia Robert Kanjirethingal

Supervisor: Nadežda Spiridovska

Degree: Professional Master

Year: 2024

Work Language: English

Study programme: Aviation Management

More...

Key factors of financial resilience in the market segment of full service carriers in crises

This thesis focuses on a detailed analysis of the financial resilience factors within the airlines of the full-service market segment. The thesis is based on exploring the factors that enable airlines to manage and recover from disruptions effectively. By examining the airline industry's response to the disruptions, the thesis shows important elements such as financial stability, operational agility, strategic foresight, and important management methodologies to survive crises. Additionally, important factors and important strategies and scenarios will be shown to survive and to be resilient within crises. Furthermore, the role of governmental aids and their methods, in the form of financial aid and flexibility, is examined and also shows the importance in lowering negative economic effects for the industry in crisis. The thesis also shows how important effective collaboration between airlines, regulatory authorities, and other stakeholders can help to manage crises effectively. Taking all this into account, the thesis is to show how financial resilience in the airline industry can be described in different ways. Moreover, it reveals the importance of strategic planning in network, operation and finance to handle further crises in future.

Author: Stefan Bonhardt

Supervisor: Jeļena Popova

Degree: Professional Master

Year: 2024

Work Language: English

Study programme: Aviation Management

More...

Development of a conceptual approach to building a marketing strategy for the oil and gas industry

Abstract This master thesis has aimed to develop conceptual approach to build marketing strategy for oil and gas industry’s company by maximizing market opportunities developing stakeholder engagement, as well as ensuring long-term sustainability.To rich the aim of the research the following research objectives have been solved in this study:● To examine current market trends, based on 4P’s, regulatory frameworks, as well as emerging opportunities based on literature review, studying best practices, and interviewing marketing management professionals.● To address untapped market segments by assessing competitor strategies to inform strategic positioning with differentiation. ● Develop a conceptual approach to build a robust 4Ps marketing strategy utilizing digital technologies to improve brand awareness as well as strengthen the loyalty of all stakeholders such as investors, customers.

Author: Vijay Singh Thakur

Supervisor: Irina Kuzmina-Merlino

Degree: Master

Year: 2024

Work Language: English

Study programme: Business and Management

More...

Application of time series algorithms for container imbalance forecasting using event data.

The research aim is to evaluate a time series models application for container imbalance forecasting using container event data. Compare the model results and conclude the real-world application options and limitations. The research object is time series models of container imbalance forecasting and subject is performance of models for container imbalance forecasting based on event data.There are three chapters of the research. The first one is State of the Art on Empty Container Repositioning (ECR) forecasting methods and approaches. The second part is investigation of container imbalance forecasting opportunities using event data. The third part is an application of time-series methods for forecasting container imbalance, experiments with real data and attempts to develop a novel data-driven framework for event data trained time series model evaluation. The third part consists of training experiment results analysis and interpretation. The 8 different models of ARIMA, VAR, VECM algorithms were tested and evaluated by different container size and type combinations, as well of 6 different port locations. Finally, the research conclusions are followed by references and attachments.

Author: Vjačeslavs Matvejevs

Supervisor: Dmitry Pavlyuk

Degree: Master

Year: 2024

Work Language: English

Study programme: Computer Sciences

More...

Table View
Text View