Sleep is a natural and reversible state of rest. In mammals it is characterized by the cyclical occurrence of Rapid Eye Movement (REM) and non-REM sleep. The latter is characterized by the presence of sleep spindles, K-Complexes (KCs), slow cortical waves and fast hippocampal waves. During sleep there is a spontaneous reactivation of recently acquired information. The slow waves orchestrate the hippocampal-cortical communication by synchronizing their activity with the fast thalamic sleep spindles and the fast hippocampal waves, favouring the transfer of information from the hippocampus to the cortex and its long-term consolidation. KCs are a type of slow wave that can be evoked by different stimuli or occur spontaneously during sleep. Studies indicate that it is possible to induce the reactivation of specific memories through the presentation of cues linked to the learned task resulting in induced KCs and memory enhancement.Different types of slow oscillatory activity (slow and delta waves) are regularly identified by visual inspection or with methods based on the morphological analysis of waveforms. In this work, we propose a method based on Machine Learning techniques, extracting features and using classification algorithms to identify signal segments. On the other hand, we differentiate between induced and spontaneous KCs. Different statistical parameters of accuracy and efficiency of the algorithm are contrasted against visual inspection provided by experts.