Machine learning approaches gain more and more importance for fatigue assessment of materials and industrial parts. In this work an extensive database of more than 22.000 single fatigue test and 1100 fatigue test series (SN-curves) of different steels are used to build a generalized approach for the fatigue prediction based on machine learning (ML). For this, different strategies are used: First, the fatigue assessment based on SN-curves, where the SN-curve parameters (slope, characteristic fatigue strength, …) were determined by ML and used for the fatigue life prediction; and second, the fatigue life prediction based on specimens, where the characteristics of single specimen of the fatigue tests series (stress amplitude, number of load cycles, …) are used. Different ML approaches (artificial neural networks, random forest, …) approaches are used. A higher accuracy of the fatigue life prediction by ML is shown for the SN-curve approach. Also, a recommendation is given by the results of this work how data should be arranged, and which characteristics (or parameter) should be used for the fatigue assessment of steels by ML.