Lameira, PedroNimtz, Julius2019-06-252022-06-302019-01-23http://hdl.handle.net/10362/73608This paper,Asset Allocation using Machine Learning, proposes a two step model, forecasting rst volatility through an GJR-GARCH model and using a Support Vec- tor machine to do the investment decision between the market portfolio and a risk parity portfolio. Besides the volatility forecast, the Support Vector Machine is based on economic, price, fundamental and sentiment data. It manages to outper- form both the market (S&P 500) and a risk parity portfolio in terms of returns and risk adjusted returns.engSupport vector machineAsset allocationRisk parityVolatility forecastmaster thesis202225917