International Research Journal of Engineering and Technology (IRJET) e-ISSN: 2395-0056
Volume: 09 Issue: 03 | Mar 2022 www.irjet.net p-ISSN: 2395-0072
© 2022, IRJET | Impact Factor value: 7.529 | ISO 9001:2008 Certified Journal | Page 1880
Bitcoin Price Prediction for Long, Short and Medium Time Frame
Shreedeep Nair
1
, Aditya Singh
2
, Rutvij Nerurkar
3
, Dr. Suvarna Pansambal
4
123
BE Student, Computer Science, Atharva College of Engineering, Mumbai, India
4
Head of Department, Computer Science, Atharva College of Engineering, Mumbai, India
---------------------------------------------------------------------***---------------------------------------------------------------------
Abstract - Over the past decade the world has seen a gradual shift in the use of digital methods than the actual methods
used in the past. The key to this economic change is the collection of digital crypto -currency. Among this group of crypto-
currencies, the most widely used in today's market is Bitcoin. Bitcoin is a blockchain-based digital asset that is not under the
control of any central authority and uses peer to peer financial performance. Due to the nature of crypto -currencies, excessive
flexibility is seen in that way it encapsulates the need to study basic value models and exhibits unstable behaviors as the
statistical distribution of data changes over time. To address these issues, we use classification based on machine learning and
retraction models. Although the machine learning-based classification models used to date have only been able to study for
one day, this paper focuses on the need to use a highly efficient machine learning model based on differentiation and retrieval
that can be read over a period of time. for one day, one week, 30 days and 90 days of time. This paper focuses not only on the
use of a highly efficient machine learning model but also on the use of a possible model, which is able to produce the correct
accuracy over a specified period of time.
Key Words: Time Series Forecasting, Blockchain, Machine learning, Deep Learning, Regression, Classification.
1. INTRODUCTION
Economics and financial systems around the world are changing to digital at an unprecedented rate, as it is believed that
digital economic integration is one of the major disruptions in the global financial system. It is estimated that by 2025 the
size of digital assets comprising both tangible and intangible assets should include 25% of the total value of 23 trillion USD.
Blockchain also appears to have found its place among similarities between Fintech and next -generation networks due to
advances in Distributed ledger (DLT) technology, which is directly responsible for the creation and use of digital assets.
Intangible property is an asset that lacks any tangible object; therefore crypto-currencies are considered intangible. It is an
issue with crypto-currency due to its fragmentary nature of extreme price fluctuations. Figure 1 shows the worst case
scenario for BTC prices during the period April 1, 2013 to December 31, 2019. Figure 1 shows a 1900% increase in BTC
prices for 2017, only followed by 72. % decline in its value in 2018. As a digital asset Bitcoin is considered a strong
indicator, although it shows extreme volatility in its values. It is also believed that BTC can quickly recover its value even if
the market is in a state of uncertainty.
Fig:1: BTC prices during the time period of April 1,2013 to December 31, 2019L
Various studies have been presented in an effort to predict BTC prices. [2] One such study attempts to predict daily price
changes by identifying different parameters and daily trends in order to gain insight into specific relevant factors. This
study uses MATLAB's neural network toolbox to build and evaluate network performance Takes normal log values as input