Matteo Chinazzi

About me

I am a 3rd year PhD student in Economics at Sant'Anna School for Advanced Studies (Pisa, Italy). My previous academic background was also in Economics, since I did both my bachelor (Economics and Management of Innovation and Technology) and master (Economics and Social Sciences) at Bocconi University in Milan.

Current Research Interests

My current research interests, in a nutshell, revolves around the idea of understanding how topological structures of financial networks influence agents' behaviors and performances in case of "critical events". In particular, I am working on two projects:

Macro-financial networks: I analyzed the data collected by the International Monetary Fund (IMF) about bilateral financial exposures between countries (e.g. cross-equity holdings or borrowing-lending relationships) and tested whether knowing the "position" of a country within the international financial network can tell us anything about the intesity of the crisis "felt" by that specific node/country. In particular I looked at stock market performances in (roughly) 50 countries. For details, here you have the working paper: Post-Mortem Examination of the International Financial Network.
Micro-financial networks (work-in-progress): I am analyzing high-frequency transactions data to understand whether it is possible to: 1) classify brokers' trading behaviors; 2) predict how brokers will respond to specific shocks (e.g. what would happen if Apple stocks collapse by 10% in one day?); 3) predict how the topology of the (bipartite) network will behave under different market conditions.

Past Research Interests

In the past, I worked with agent-based simulations and models of epidemic diffusion. In particular, I helped building a model for creating synthetic social mixing matrices of the Italian population that were then used to estimate a variety of age-specific contact matrices. Those matrices were then validated against Italian serological data for Varicella (VZV) and ParvoVirus (B19) in order to see whether our simulations were able to capture the same transmission patterns of the diseases as observed in real data (at least from a qualitative point of view). If you want to know how successful we were, take a look at the published paper here: Little Italy: An Agent-Based Approach to the Estimation of Contact Patterns- Fitting Predicted Matrices to Serological Data.

Future Research Interests (???)

Future projects are still in a state of flux but - just to give you an idea of the topics range - those are the sorts of data I am collecting:

Soccer data: e.g. detailed match statistics for different national leagues and international tournaments, players economic valuations over time, data about players' transfers, etc..
Twitter data: i.e. "tweets networks"
Music,Movies,TV series and Videogames data: e.g. user/viewer reviews, launch dates, ratings, box office revenues, etc..
App data: e.g. iOS and Play store data on app downloads, user ratings, prices, etc..

Expertise

Over the years, I managed to work with many different programming languages and I used many scientific software packages. As far as programming languages are concerned, I did my master thesis in Java and C, while at the moment I am mainly working in Python (which became my favorite programming language by far [unless I have to resort to C for computational reasons.. ;) ]).

In terms of software packages, I used the following (ordered by proficiency): Matlab, Octave, Gretl (econometric software), STATA, R, SPSS, XLSTAT (addon for Excel) and Eviews.

I also work with many software libraries:

in Python: networkx, numpy, scipy, pylab, mysqld, sqlite3, json, urllib, re, etc..
in C: GSL
in Java: Repast

Free time

Before grad school (aka "when I really had free time to spare"), I used to ski and play basketball quite a lot, I was a semi-professional bowler and I had fun playing tennis table and billiards. In addition, being Italian, I also enjoyed play soccer.