The recognition of modulation schemes in military and civilian applications is a major task for intelligent receiving systems. Having no previous knowledge of the transmitted signal as well as uncertainties in the channel and the receiver makes the identification of the modulation scheme a difficult task. Various Automatic Modulation Classification (AMC) algorithms have been developed to overcome this challenging task. However, classification with low computational complexity as well as reasonable processing time is still a challenge, especially, for modulation types with similar constellations under realistic channel conditions. In this paper, a feature-based approach along with various classifiers is employed based on statistical features as well as higher-order moments and cumulants. First, some well-known classifiers including Decision Trees (DT), Support Vector Machines (SVM), K-Nearest Neighbors (KNN) and ensemble are evaluated at different SNR values. Then, various forms of SVMs have been utilized. An over-the-air (OTA) recorded dataset consisting of four analog and ten digital modulation schemes are used for testing the proposed method. The classification performance is compared against SNR (0-20 dB). The overall accuracy for quadratic SVM is found to be as high as 98% at 10 dB. This has been comparable with the reported performance based on simulated dataset. Overall, the comparison of the results with those published in the literature indicates that this is the first paper presenting such a high accuracy with OTA dataset consisting of fourteen modulation schemes.