The study aims to augment commonly applied volatility models with support vector machines and neural networks. Further, fractional integration and asymmetric powers will be introduced. The proposed modeling strategy benefits from neural network based GARCH models and SVR-GARCH models. Following these approaches, the study proposed fractional integration and asymmetric power GARCH structures to obtain SVR-FIAPGARCH and NN-FIAPGARCH models to be evaluated in terms of learning algorithms. Models are evaluated for in-sample and out-of-sample forecasting of daily returns in Istanbul ISE100 stock index. Results suggest several findings: i. fractional integration and asymmetric power structures could be modeled with learning algorithms. ii. volatility clustering, asymmetry and nonlinearity characteristics are modeled more effectively with SVR-GARCH and MLP-GARCH models compared to the GARCH models. iii. SVR-GARCH models provided the lowest error criteria levels in out-of-sample and are closely followed by the MLP-GARCH models.