Machine Learning approaches in Neuroscience:behavioral and sleep classification

The understanding of sleep is of paramount importance from a scientific and clinical point of view. Brain disorders, such as Autism and Alzheimer, show disrupted sleep patterns that contribute to the progression of the disease. To obtain efficacious drugs, an important step is to study and test them in rodents, with the hope to extrapolate the findings to humans. In such preclinical studies, it is fundamental to correctly identify and classify sleep phases in order to compare them to the ones found in humans. However, very few works have been carried out in this regard. This Master thesis work is aimed at critically studying this aspect through an approach based on Machine Learning techniques applied to EEG and accelerometer signals. This work will lay the foundation to investigate the differences between wildtype and transgenic mice with the purpose of characterizing the sleep impairment biomarkers of the disease and its trajectory throughout the rodent's life.