CNN-BASED INVESTMENT STRATEGY USING TECHNICAL INDICATORS

Abstract

Since the dawn of time, both individuals and institutions have been in pursuit of strategies to amplify their wealth, recently, more in the financial world. This process of increasing one’s wealth changed and evolved in the same manner as did technological progress, particularly with the integration of artificial intelligence (AI) and machine learning (ML) technologies in the financial analysis. Among these advancements and innovations, Convolutional Neural Networks (CNNs) have emerged as a formidable tool for forecasting market trends and predicting outcomes of transactions. Current investment strategies mostly use numerical data, technical indicators derived from numerical data, and the intricate algorithms of neural networks (NNs). However, the current methodologies for data transformation misses the necessary diversity to fully exploit the capabilities of advanced neural network models. This paper aims to bridge this gap by proposing a novel approach that utilizes models with technical indicators with new data format. The end goal of this paper is to develop a robust investment strategy that utilizes power of NNs while using uncom- mon data transformation in order to achieve great results in accuracy and reliability of trend prediction. To achieve this goal, this paper introduces a methodological know-how that methods that involve converting technical indicators into images and feeding them to NNs, specifically NNs that are well adept in image classifications. Particularly, leveraging CNN’s excellence in detecting unseen patterns within these visual representations, offering a enhancing the effectiveness of investment strategies through the power of image-based data analysis.