Department of Economics Speaker Series

Thursday, March 20, 3:30-4:45 p.m.

EN-2006

The Department of Economics is pleased to welcome Dr. Federico Severino, who will speak about current machine learning applications in business administration.

Financial asset prices display recurrent patterns over time but such time series are usually noisy and volatile, making the identification of repetitive patterns difficult.

“Functional motif discovery in stock market prices” (joint with Marzia A. Cremona and Lyubov Doroshenko) embeds asset prices in a functional data analysis framework, by extending and using probabilistic K-means with local alignment to discover functional motifs in stock price time series. It then exploits information of the discovered motifs to perform the price forecasts with a novel motif-based algorithm. The technique is illustrated on simulations of mixed causal-noncausal autoregressive processes and apply it to the prices of S&P 500 top components.

Dr. Severino received his Ph.D. in Economics and Finance from Università Bocconi. He is an Associate Professor at Université Laval. He is also a collaborating researcher of the Institute Intelligence and Data, a researcher of CIRANO and a researcher of the Financial Engineering Laboratory at Université Laval.

Presented by Faculty of Humanities and Social Sciences

Event Listing 2025-03-20 15:30:00 2025-03-20 16:45:00 America/St_Johns Department of Economics Speaker Series The Department of Economics is pleased to welcome Dr. Federico Severino, who will speak about current machine learning applications in business administration. Financial asset prices display recurrent patterns over time but such time series are usually noisy and volatile, making the identification of repetitive patterns difficult. “Functional motif discovery in stock market prices” (joint with Marzia A. Cremona and Lyubov Doroshenko) embeds asset prices in a functional data analysis framework, by extending and using probabilistic K-means with local alignment to discover functional motifs in stock price time series. It then exploits information of the discovered motifs to perform the price forecasts with a novel motif-based algorithm. The technique is illustrated on simulations of mixed causal-noncausal autoregressive processes and apply it to the prices of S&P 500 top components. Dr. Severino received his Ph.D. in Economics and Finance from Università Bocconi. He is an Associate Professor at Université Laval. He is also a collaborating researcher of the Institute Intelligence and Data, a researcher of CIRANO and a researcher of the Financial Engineering Laboratory at Université Laval. EN-2006 Faculty of Humanities and Social Sciences