Cluster-Based Mutual Fund Classification and Price Prediction Using Machine Learning for Robo-Advisors.

Xiaofei Chen,Shujun Ye,Chao Huang,José Alfredo Hernández-Pérez

doi:10.1155/2021/4984265

Abstract

The rise of FinTech has been meteoric in China. Investing in mutual funds through robo-advisor has become a new innovation in the wealth management industry. In recent years, machine learning, especially deep learning, has been widely used in the financial industry to solve financial problems. This paper aims to improve the accuracy and timeliness of fund classification through the use of machine learning algorithms, that is, Gaussian hybrid clustering algorithm. At the same time, a deep learning-based prediction model is implemented to predict the price movement of fund classes based on the classification results. Fund classification carried out using 3,625 Chinese mutual funds shows both accurate and efficient results. The cluster-based spatiotemporal ensemble deep learning module shows better prediction accuracy than baseline models with only access to limited data samples. The main contribution of this paper is to provide a new approach to fund classification and price movement prediction to support the decision-making of the next generation robo-advisor assisted by artificial intelligence.

Highlights

In recent years, machine learning and deep learning have been widely used in finance to meet financial needs [1–5]
As robo-advisors are affordable to common investors, these advisors have become one of the main tools for financial institutions to carry out wealth management innovation for several advantages, e.g., they can manage long-term asset allocation, applying modern portfolio theory with technologies, such as big data, and implementing cloud computing. ey can automatically provide clients with fund investment suggestions through the Internet, with the consideration of investors’ risk preferences, property status, and financial objectives
We cluster the fund via Gaussian mixture model (GMM) using fund features described in the previous section, from which the dimension of the fund features is reduced using the principal component analysis (PCA) method. en, we develop a deep learning-based prediction model to predict the fund trend based on the clustering results. erefore, we will briefly introduce the PCA and GMM at first. e deep learning-based model is presented in detail

Summary

Introduction

Machine learning and deep learning have been widely used in finance to meet financial needs [1–5]. Based on the GMM results, we use the spatiotemporal ensemble deep learning model to predict the short-term price movement of each category of funds. To compare the performance of predicting fund net asset values of our model, we compare it with several basic models, that is, residual network (ResNet, hereafter) model, long- and short-term memory network (LSTM, hereafter) model, and one-dimensional convolutional neural network (CNN, hereafter) model We examine their performance in predicting the short-term returns of the four main classified categories in our results by employing the mean absolute error (MAE) and correlation coefficient R2 as evaluation indicators. E main contributions of this paper are as follows: (1) we propose a new two-step GMM model to effectively distinguish the mutual funds in China using simple fund characteristics and (2) we construct an ensemble deep learning model to predict the short-term price movement of different categories of funds

Literature Review

The Model

Results

Deep Learning-Based Prediction Model