在近幾年，有很多機器學習的技術演進，而風險溢酬在市場一直都是重要的議題。本篇文章整理一些技術的歷史，並模擬市場上的資產配置和投資策略，嘗試用強化學習捕捉市場行為，並透過添加風險溢酬與否以及現代投資組合理論，發展風險溢酬在市場的重要性的依據。簡短來說，我們模擬了宏觀經濟的行為然後計算風險偏好和獎勵值，在策略不變的情況下去找到風險溢酬的重要性。;Machine Learning techniques improve a lot in recent years. At the same time， risk premia are important issue in the market for a long time. This article organize some of the history of those techniques， and simulate allocation and strategies in the market， trying to capture market behavior using Reinforcement Learning. We also compare whether adding risk premium or not and Modern Portfolio Theory to develop evidence of significance of risk premia in the market. In short， we model the macroeconomics and their behavior， and calculate risk preference and reward without any change of strategy， in order to find out if risk premium has the importance.
