CincoCrypto - A Cryptocurrency Price Forecasting Tool For Everyone
Author
Lara de Leyva, Gonzalo
Term
4. semester
Education
Publication year
2023
Submitted on
2023-05-21
Pages
75
Abstract
Cryptocurrencies have grown rapidly over the past decade, and so have investment scams. To support safer, more informed decisions, this project introduces CincoCrypto, a simple, open-source price forecasting tool for all users. CincoCrypto trains four machine learning models (computer methods that learn patterns from data) on one year of market data to predict the next day’s price. In tests, the Space Vector Regression model was the fastest, while the Convolutional Neural Network was the most accurate of the four.
Kryptovalutaer er blevet langt mere udbredte det seneste årti, og det samme gælder investeringssvindel. For at støtte sikrere og mere oplyste beslutninger præsenterer dette projekt CincoCrypto, et enkelt, open-source værktøj til prisprognoser for alle brugere. CincoCrypto træner fire maskinlæringsmodeller (computerbaserede metoder, der lærer mønstre fra data) på ét års markedsdata for at forudsige næste dags pris. I test var Space Vector Regression-modellen hurtigst, mens Convolutional Neural Network var mest præcis af de fire.
[This apstract has been rewritten with the help of AI based on the project's original abstract]