The Machine Learning Race: Can Algorithms Truly Predict Bitcoin’s Volatility
Date
2-20-2025
Faculty Mentor
Monica Trifas, Mathematics, Computing & Information Sciences
Files
Submission Type
Conference Proceeding
Location
8:15-8:25 am | Houston Cole Library, 11th Floor
Description
An unpredictable, chaotic, and endlessly fascinating whirlwind is the Bitcoin market. We test the limits of human intelligence and machine accuracy as we observe its erratic movements, posing the straightforward but important question: Can algorithms really make sense of something so unpredictable?
We tested a variety of machine learning models in this study, ranging from more sophisticated neural networks like RNNs and LSTMs to more conventional techniques like decision trees and support vector machines. Like a lone explorer, each model looks for the mysterious patterns concealed in the price and volume swings of Bitcoin. While deeper models take their time, processing the complexities, but at a cost, others move quickly but only touch the surface.
But which strategy is most effective? That's what we want to learn. By contrasting these models side by side, we can determine whether more straightforward algorithms are sufficient or whether deep learning's added complexity actually gives an advantage. Does efficiency and speed trump more profound understanding? Or do complexity and patience result in more accurate forecasts?
However, this is about more than technology. It speaks to something deeply human—the insatiable desire to predict, comprehend, and impose order on chaos. Regardless of how sophisticated our models become, Bitcoin remains unpredictable, reminding us that even the brightest minds and most advanced algorithms are sometimes simply looking for meaning in the unknown. Some modern neural networks detect many noises in data, whereas traditional algorithms barely detect anything. The approach to find the perfect model should be by utilizing the different algorithms and adding the market sentiments through social media, news, and many more.
Keywords
student research, computing
Rights
This content is the property of Jacksonville State University and is intended for non-commercial use. Video and images may be copied for personal use, research, teaching or any "fair use" as defined by copyright law. Users are asked to acknowledge Jacksonville State University. For more information, please contact digitalcommons@jsu.edu.
Disciplines
Computer Sciences
Recommended Citation
Basyal, Prashant, "The Machine Learning Race: Can Algorithms Truly Predict Bitcoin’s Volatility" (2025). JSU Student Symposium 2025. 15.
https://digitalcommons.jsu.edu/ce_jsustudentsymp_2025/15
