NeuroFi: An Intelligent Multi-Model AI Platform for Crypto Price Forecasting and Smart Trading
Summary
No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.
Summarise with VIRAT Free to try with your Science Catalogs account.Abstract
The cryptocurrency market exhibits extreme volatility and non-linear price dynamics, presenting significant challenges for accurate price prediction and trading decision support. Existing approaches often rely on single-model architectures that fail to capture the multifaceted nature of market movements influenced by technical patterns, market sentiment, and temporal dependencies. This paper presents NeuroFi, a comprehensive AI-powered trading decision support framework that integrates multiple prediction models, real-time sentiment analysis, and technical indicators through a novel ensemble architecture. The proposed system combines Long Short-Term Memory (LSTM) deep learning networks with statistical models including linear regression, moving average convergence-divergence (MACD), and Relative Strength Index (RSI) analysis. Additionally, NeuroFi incorporates a dual-engine sentiment analysis pipeline utilizing VADER and TextBlob algorithms to process news and social media data from multiple sources. The framework introduces a risk-aware recommendation engine that generates personalized trading signals based on configurable risk profiles. Implemented using a three-tier architecture with React, Express.js, and FastAPI, NeuroFi demonstrates the feasibility of real-time multi-model integration for cryptocurrency trading applications. The proposed methodology addresses critical gaps in existing literature by providing adaptive ensemble weighting, multi-source data fusion, and explainable trading recommendations.
Fields of Study & Subjects
Stock Market Forecasting Methods
Blockchain Technology Applications and Security
Financial Markets and Investment Strategies
Publication details
- Journal / source
- International Journal for Research in Applied Science and Engineering Technology
- Volume
- 14
- Issue
- 3
- Pages
- 4657–4664
- Publication date
- 31 March 2026
- Language
- EN