Stall Induced Vibrations (SIV) are an aeroelastic instability that might happen when large portions of the wind turbine blade operate in moderate stall, which leads to high internal loads that can damage the structure. Exploring the inflow space that cause SIV has a high computational cost, which quickly increases with the number of input variables, since it involves high-fidelity aeroelastic simulations. In this work, a surrogate model is used to effectively explore the behavior of SIV in a five variable inflow space consisting of wind speed, yaw angle, vertical wind shear, wind veer, and atmospheric temperature. The surrogate model, trained from the results of a few aeroelastic simulations, predicts the severity of SIV for a given inflow condition. The surrogate model is chosen by comparing the performance of a few common model types, and Artificial Neural Networks (ANN) are found to provide a good balance between accuracy and computational time. Using the ANN, the occurrence and severity of SIV in the inflow space is studied. The analysis shows that for the turbine studied, yaw angle is the most influential variable, while temperature is the least influential variable in terms of conditions that lead to occurrence of SIV. Inflow conditions with a moderate yaw angle (around 10-25 deg), high wind speeds, and high negative veer are found to be the most critical in terms of severity of SIV. The current work is expected to serve as a helping tool to decide the focus of computationally expensive simulations such as high fidelity CFD based aeroelastic simulations.