Back to search

Institutional Repository of Vilnius University

Akcijų kainos prognozavimas naudojant sentimentų analizę /

Abstract

dc:description

Prediction is used in a lot of areas, for example, in finance, business or weather forecastig, but it is certainly not a trivial task. Being able to predict stock price accurately would greatly help every investor make informed decisions, which would increase his/her profits. There is a widespread opinion that stock prices follow a random-walk pattern and cannot be predicted. However, with the emergence of social media, and machine learning, more researchers believe that public sentiment can indeed help predict stock prices. This work discusses stock price prediction using sentiment analysis. At first, sentiment analysis methods are reviewed and an algorithm chosen to predict stock prices, XGBoost is investigated further. Then an experiment was carried out in order to understand if sentiment analysis, extracted from one of the most popular social media platforms, Twitter, can lead to better results while predicting stock prices. With the help of parameter tuning, and choosing an appropriate day window to predict the stock price, the results look promising: a 2,924 % decrease in RMSE is observed when predicting stock prices for 1 day into the future when using sentiment scores additionally to close prices.

Degree

thesis:*
Grantor dc:publisher
Institutional Repository of Vilnius University
Year dc:date
2022

Author and committee

dc:creator, dc:contributor.*
Author dc:creator
  • Sabaliauskaitė, Gintarė,

Rights

dc:rights
Statement dc:rights
  • info:eu-repo/semantics/openAccess
Language dc:language
lit

Identifiers

dc:identifier.*

Chain of custody

source
Harvested from
Vilnius University
Base URL
epublications.vu.lt/oai
Last updated
2026-11-19
Source record
OAI-PMH GetRecord
related terms
citation

Sabaliauskaitė, Gintarė,. Akcijų kainos prognozavimas naudojant sentimentų analizę /. Institutional Repository of Vilnius University, 2022. https://repository.vu.lt/VU:ELABAETD146237075&prefLang=en_US