ADS Capstone Chronicles Revised

3

2.2 Definition of Objectives The “Guardians of the Crypto” project aims to develop a holistic application with a primary function: predicting future cryptocurrency price changes based on historical data. This project aims to create a predictive model using historical price data to forecast future price movements. These tools will be integrated into a user-friendly application, allowing traders and investors to access predictive price forecasts. The effectiveness of these tools will be validated to improve decision-making accuracy and reduce financial risks. By achieving these objectives, a comprehensive tool will be provided to help traders and investors make more informed decisions. If the answers are validated, users will have a reliable means of predicting market movements, leading to better investment strategies. If the answers are disproved, insights into the limitations of current analytical techniques will be gained, and alternative approaches to address the identified problem will be explored. Bhatt et al. presented an innovative study on cryptocurrency price prediction. Leveraging historical price data and machine learning techniques, including neural networks, their approach modeled how market sentiments from social media can influence cryptocurrency pricing. The research is grounded in the hypothesis that when integrated with traditional price data, social sentiment indicators have the potential to provide a more robust and accurate framework for predicting cryptocurrency price movements, offering a promising future for the field. A significant strength of the study is its comprehensive use of data sources, blending quantitative price data to enhance prediction accuracy. The study meticulously and rigorously 3 Literature Review 3.1 Bhatt et al. (2023)

evaluates various machine learning models to determine which best captures the complexities of the market influences, providing a robust framework for cryptocurrency price prediction. However, Bhatt et al. ’ s research also encounters some limitations. The primary concern is the challenge of real-time data processing and the scalability of their model across different cryptocurrencies beyond the primary ones like Bitcoin and Ethereum. Additionally, while the paper thoroughly investigates the correlation between sentiment and price, it somewhat overlooks the potential latency or lag effect of sentiment on market prices, which could be crucial for real-time trading strategies. In their groundbreaking study, Bollen et al. investigated the predictive power of Twitter moods on stock market trends. The authors used sentiment analysis and complex algorithms to convert Twitter data into mood indicators that correlate with the Dow Jones Industrial Average. Their findings intriguingly suggested specific mood dimensions, particularly calmness, possess a significant predictive relationship with stock market movements leading up to three days in advance. Despite these compelling findings, the study ’ s scope is limited to traditional financial markets. It does not extend to the cryptocurrency sector, known for its distinct dynamics and market participant behaviors. Furthermore, while the study underscores the potential of social media sentiment in forecasting market trends, it needs to delve into the operational challenges of real-time data processing and the implications of varying data quality on predictive outcomes. 3.2 Bollen et al. (2011)

70