Time series prediction
The most standard test for a theoretical model is trying to predict new data points in a time series. This general task is ubiquitous and allows different theories to be compared quantitatively. Recently, the use of deep machine learning models has shown state-of-art performances in many repetitive, yet complex, tasks, such as image classification, speech recognition, and autonomous driving. Nevertheless, simple autoregressive models still seem to be more efficient in the context of financial time series prediction. I’m interested in developing optimal architectures for time series prediction with deep machine learning models, and in characterising their efficiency for time series from different sources, ranging from finance to physics.
Theory of intelligence
As advanced algorithms perform more and more complex tasks, the question of what intelligence really is becoming increasingly pivotal for understanding new software technology, for envisioning new directions of scientific research, and imagining our role as humans in a machine-supported society. It is a question that can be addressed from many points of view; from a machine intelligence perspective: what is an intelligent machine? How intelligent a machine is? From a biological perspective: what is an intelligent organism? From an evolutionary perspective: how does intelligence emerge and grow? What is the future of intelligence? Is there a collective notion of intelligence? All these perspectives require both a quantitative statistical analysis and a qualitative philosophical investigation, that I try to undertake.
Cybersecurity has been a central topic since the very beginning of the Internet era. Now, blockchain technology and, in particular the Bitcoin protocol, has demonstrated the potential of new security algorithms to substitute old paradigms of intermediation. A system based on a decentralised, possibly open, protocol has to regard with further attention on the role played by the properties of the network of communication, e.g. speed of information transmission, heterogeneity of connections, clustered sets of nodes. I’m interested in understanding what are the limitations and possibilities that are unlocked by the blockchain technology and especially by the role of the network of communication between the nodes of a decentralised system.
Systemic risk and macroprudential regulation
Interconnectedness in the world economy poses major challenges in monitoring financial fluxes and risks, previously unrelated financial institutions may hold similar instruments and be subject to similar risks that may propagate in their own networks of financial counterparties. Modeling systemic risk I try to make sense of structural data on financial interconnections and portfolio overlaps, considering probable economic scenarios and deriving possible patterns of financial risk propagation at a national and international level.