In this paper, we discuss an application of machine-learning-based methods to the identification of candidate active galactic nucleus (AGN) from optical survey data and to the automatic classification ofAGNs in broad classes. We applied four different machine-learning algorithms, namely the Multi Layer Perceptron, trained, respectively, with the Conjugate Gradient, the Scaled Conjugate Gradient, the Quasi Newton learning rules and the Support Vector Machines, Q4 to tackle the problem of the classification of emission line galaxies in different classes, mainly AGNs versus non-AGNs, obtained using optical photometry in place of the diagnostics based on line intensity ratios which are classically used in the literature. Using the same photometric features, we discuss also the behaviour of the classifiers on finer AGN classification tasks, namely Seyfert I versus Seyfert II, and Seyfert versus LINER. Furthermore, we describe the algorithms employed, the samples of spectroscopically classified galaxies used to train the algorithms, the procedure followed to select the photometric parameters and the performances of our methods in terms of multiple statistical indicators. The results of the experiments show that the application of self-adaptive data mining algorithms trained on spectroscopic data sets and applied to carefully chosen photometric parameters represents a viable alternative to the classical methods that employ time-consuming spectroscopic observations.