anomaly detection

Anomalies are rare and anomaly detection is often therefore framed as One-Class Classification (OCC), i.e. trained solely on normalcy. …

The progress in modelling time series and, more generally, sequences of structured data has recently revamped research in anomaly …

Considering the increasing aging of the population, multi-device monitoring of the activities of daily living (ADL) of older people …

In this presentation, I presented future directions of explainable anomaly detection in daily routine behaviours of older patients.

Online learning environments (OLEs) have seen a continuous increase over the past decade and a sudden surge in the last year, due to …

Autoencoders have been recently applied to outlier detection. However, neural networks are known to be vulnerable to overfitting, and …

Regional project - Avviso Pubblico “Emergenza Coronavirus e oltre” - Domanda prot. n. A0376-2020-070051, CUP; F84E21000000006) in support to DATAWIZARD SRL

In this tutorial, we explored challenges in retention, and presented machine learning solutions to the difficulties in predicting student dropout rates from online courses.

An Italian presentation about the current research projects of the lab. As research project for the “AI for Education” workshop, I presented the state-of-the-art in EDM, and proposed to solve the student dropout problem with Hierarchical Attention Networks (HANs).

BAE is an unsupervised ensemble method that builds an adaptive cascade of autoencoders to achieve improved and robust results. BAE trains the autoencoder components sequentially by performing a weighted sampling of the data, aimed at reducing the amount of outliers used during training, and at injecting diversity in the ensemble.