Artificial intelligence/machine learning for wealth management on mobile device

Traditionally, portfolio management involves balancing a portfolio with different assets using statistical methods of analysis. These analyses are typically performed by portfolio managers or expert investors. For the amateur investor, the level of research required to form a solid understanding...

Full description

Saved in:
Bibliographic Details
Main Author: Sim, Eccles Jia Xuen
Other Authors: Ng Wee Keong
Format: Final Year Project
Language:English
Published: Nanyang Technological University 2022
Subjects:
Online Access:https://hdl.handle.net/10356/156611
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
Description
Summary:Traditionally, portfolio management involves balancing a portfolio with different assets using statistical methods of analysis. These analyses are typically performed by portfolio managers or expert investors. For the amateur investor, the level of research required to form a solid understanding of assets can be unmanageable. In the absence of time, tools, or level of information to match the experts, this project explores artificial intelligence solutions that may aid in reducing the analytical gap between amateur investors and financial experts. Our goal is to create an application that is intuitive to an amateur investor while maintaining the technicalities required for deep valuations of portfolio assets. Apart from the ability to learn and predict optimal allocations of portfolios, the application provides supplementary features automating the analysis of a portfolio using standard modern portfolio theory (MPT) frameworks. The mobile application is developed using the Dart programming language along with the Flutter Framework. A variant of the deep reinforcement learning algorithm known as proximal policy optimization (PPO) is used as the agent to learn an investor’s portfolio and suggest optimal stock allocations for maximized returns. It is imperative to note that this mobile application is a proof of concept and is not financial advice. Keywords: Analysis; Reinforcement Learning; Mobile Application; Amateur Investor