Date of Publication
12-19-2022
Document Type
Bachelor's Thesis
Degree Name
Bachelor of Science in Mathematics with Specialization in Computer Applications
Subject Categories
Mathematics
College
College of Science
Department/Unit
Mathematics and Statistics Department
Thesis Advisor
Kristine Joy E. Carpio
Defense Panel Chair
Jose Tristan F. Reyes
Defense Panel Member
Sonia Y. Tan
Abstract/Summary
Machine learning is a method of data analysis which can be used to predict outcomes given previous and present data. The rise in the popularity of cryptocurrencies due to its volatile movements has piqued the interest of researchers and investors alike. Different machine learning models such as support vector machine, random forest, and regressors were used to predict the classification of prices by either increased or decreased and to predict the market capitalization (marketcap) of both Bitcoin and Ethereum. The machine learning models that were used as machine learning classifiers are the support vector machine and random forest. These machine learning models, and its respective algorithms categorized the given sets of data into different set of clusters or classes. In this study, machine learning classifiers are used to either label a certain data whether the price of Bitcoin and Ethereum increased or decreased based on its opening and closing value of the given day. The input data used was 187 days before the predicted price movement. To predict the exact value of the marketcap of Bitcoin and Ethereum, the linear regression model was used in the study. For machine learning classifiers, the model with the highest accuracy for both Bitcoin and Ethereum was the support vector machine linear kernel with an accuracy of 90%. For the linear regression model used to predict the marketcap of both Bitcoin and Ethereum, Bitcoin has the higher accuracy score of 94% compared to Ethereum’s 86% accuracy score.
Abstract Format
html
Language
English
Format
Electronic
Physical Description
[ii], 43 leaves
Keywords
Cryptocurrencies; Machine learning
Recommended Citation
Calceta, E. V., & Datinginoo, J. R. (2022). Testing the predictive ability of machine learning models for long term investments in cryptocurrency. Retrieved from https://animorepository.dlsu.edu.ph/etdb_math/21
Upload Full Text
wf_yes
Embargo Period
12-18-2022
