Sentiment-Driven Price Prediction of the Bitcoin based on Statistical and Deep Learning Approaches

Serafini, Giulia; Yi, Ping; Zhang, Qingquan; Brambilla, Marco; Wang, Jiayue; Hu, Yiwei; Li, Beibei

doi:10.1109/ijcnn48605.2020.9206704

Cited by 44 publications

(33 citation statements)

References 26 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Linear Regression, RNN, and LSTM [19], [20], [22], [23], ARIMAX [21], Linear discriminant analysis [24], ARIMA [25], Logistic regression, Naive Bayes, SVM [26], Multiple linear regression [27], Tweet corpus in COVID-19 era [28], Random Forest, Decision tree, AdaBoost [29], XGBoost-Composite model [30], Q-Learning [This paper].…”

Section: Striving For Accurate Predictionmentioning

confidence: 99%

See 1 more Smart Citation

Twitter Attribute Classification with Q-Learning on Bitcoin Price Prediction

Sattarov¹,

Choi²

2022

Preprint

View full text Add to dashboard Cite

Aspiring to achieve an accurate Bitcoin price prediction based on people's opinions on Twitter usually requires millions of tweets, using different text mining techniques (preprocessing, tokenization, stemming, stop word removal), and developing a machine learning model to perform the prediction. These attempts lead to the employment of a significant amount of computer power, central processing unit (CPU) utilization, random-access memory (RAM) usage, and time. To address this issue, in this paper, we consider a classification of tweet attributes that effects on price changes and computer resource usage levels while obtaining an accurate price prediction. To classify tweet attributes having a high effect on price movement, we collect all Bitcoin-related tweets posted in a certain period and divide them into four categories based on the following tweet attributes: (i ) the number of followers of the tweet poster, (i i ) the number of comments on the tweet, (i i i ) the number of likes, and (i v ) the number of retweets. We separately train and test by using the Q-learning model with the above four categorized sets of tweets and find the best accurate prediction among them. Especially, we design several reward functions to improve the prediction accuracy of the Q-leaning. We compare our approach with a classic approach where all Bitcoin-related tweets are used as input data for the model, by analyzing the CPU workloads, RAM usage, memory, time, and prediction accuracy. The results show that tweets posted by users with the most followers have the most influence on a future price, and their utilization leads to spending 80% less time, 88.8% less CPU consumption, and 12.5% more accurate predictions compared with the classic approach.

show abstract

Section: Striving For Accurate Predictionmentioning

confidence: 99%

“…al. [21] compared two models used for Bitcoin time-series predictions: the Auto-Regressive Integrated Moving Average with eXogenous input (ARIMAX) and RNN. The flow of studies that adopted LSTM to make a price prediction has been continued by Ye et.…”

Section: B Striving For Accurate Predictionmentioning

confidence: 99%

Twitter Attribute Classification with Q-Learning on Bitcoin Price Prediction

Sattarov¹,

Choi²

2022

Preprint

View full text Add to dashboard Cite

show abstract

“…This type of analysis can be useful for long term price sentiments, whereas traders needed more specific analysis, which is much suitable on price prediction and price forecasts. The work [4] represents, such a sentiment analysis work, where the data fed from crowd sourced data and economic data. The crowd sourced data in terms collected from twitter data.…”

Section: Related Workmentioning

confidence: 99%

Crypto-Currency Price Prediction using CNN and LSTM Models

Mounika¹

2021

IJRASET

View full text Add to dashboard Cite

Growing demand in internet usage has bring up to the revolution in digital economy. From the past decade digital economy is playing major role round the world. The physical assets are converted to digital assets, in which Bitcoin is the major crypto currency for the investors. Bitcoin has the biggest market capital among all other crypto currencies. Bitocoin price prediction and forecast has been one of the tedious tasks for long period. Existing works are attempted to go for accurate prediction and forecast through machine learning models. The forecast can be useful work for the investors to know about the nature of price in future and gives them directions for their investments. In this proposed work, Bitcoin price prediction is proposed through the deep learning models such as Convolutional Neural Networks (CNN) and Long short term memory (LSTM) models. The aim of the work is to give accurate predictions and forecast and bring the daily trend for crypto currency market. Experimental results shows that the proposed system given better accuracy on predictions.

show abstract

“…Therefore, by trying to detect all types of manipulation with machine learning and other methods used in the field of artificial intelligence, it is to reveal how effective the manipulators behind all these events can be in the detection phase. Machine learning, especially deep learning, has been used in multiple fields and industries [2,3,4]. Additionally, we achieved some experiments about manipulations happened in economically crisis periods which might be the reason of price changes on stock markets.…”

Section: Introductionmentioning

confidence: 99%