With the rising popularity of Software Defined Radios (SDR), there is a strong demand for automatic detection of the modulation type and signal parameters. Automatic modulation classification is an approach to identify the modulation type and its parameters such as the carrier frequency or symbol rate. In electronic warfare, it enables real-time signal interception and processing. In civil applications, it can be used, e.g., by the amateur radio operators to automatically set the transceiver to the appropriate modulation and communication protocol. This paper presents a modulation classification driven by a neural network. A set of signal features are provided as an input of the neural network. The paper discusses the relevance of different signal features and its impact on the success rate of the neural network classification. The proposed approach is tested on both artificial and real samples captured by the SDR.