ML Milan is back in February and we will do a deep dive into time series.
Events will be held in English with our classical format at Politecnico.
PS: due to the high number of requests to attend the event, we moved the event in the same building in a bigger room, Giulio Natta
Talk 1: “Approaches to detect anomalies in multivariate time series data”
Anomaly detection has many applications ranging from fraud analysis to detecting malware or spotting anomalous trends in financial data. The goal of this talk is twofold: (i) to provide a common definition of outlier (global, contextual, collective), and (ii) to present, by using a case study, approaches that can be used to detect and rank anomalies.
Fabio Fumarola, Prometeia
Fabio works in Prometeia as machine learning engineer to contribute in changing the way banks and insurances used data to create innovative services. Previously, he worked in the Italian Digital and Transformation Team, Unicredit R&D. As academic, Fabio received a Ph.D. in Data Mining and Machine Learning in May 2011, worked at the University of Illinois, at University of Bari, Policlinic Hospital in Bari.
Talk 2: “Finding the unexpected: A time series forecasting use case”
Being able to track your business and drive it forward is essential. In this talk, we will go through an approach for analyzing video streaming data to find answers to these questions: which KPIs influence the streaming performance? How to detect anomalous patterns that are becoming the new norm? Alternatively, how to trace the root causes of anomalies?
Golnazsadat Zargarian, Altran
Golnaz works as a consultant and engineer in R&D sector of Innovation and Strategy at Altran. Currently, she is developing machine learning-based approaches for predictive analytics.
Golnaz holds a master degree in Communications and Computer Networks Engineering from Politecnico di Torino. During her thesis, she created a semi-automated predictive model for TIM.
Where? Join us on Thursday 13th of February at Politecnico di Milano, in Piazza Leonardo.