Optimal Reconstruction of Complex Networks

The research activities will be the core of the project "Optimal Reconstruction of Complex Networks", which aims at developing an algorithmic framework for the optimal reconstruction of networks when the only available information is the in/out node strengths and, possibly, the in/out node degrees. The core of the project will be the improvement of the fitness-induced exponential random graph (fERG) method and the efficient implementation of the algorithm.
 
The selected candidate will be required to use knowledge about numerical optimization to improve the numerical reliability of the state of the art algorithm: the particular structure of the equations to be solved typically leads to instability issues which will require ad-hoc reformulations in order to ensure reliability, efficiency, and accuracy of the algorithm.
 
Furthermore, a good programming background is required: the candidate will implement the algorithm in order to deliver a tool which can be used by the scientific community. To that end, the software shall be interfaced to a high-level programming language such as, e.g., Python, be easy to install, user-friendly and well documented.
 
Finally, knowledge in economics will be required to apply the developed tool to the analysis of practically relevant case studies. Two applications will be targeted:
- social networks in villages
- firm-level production networks
For both applications, real data is available to the team, which makes the study of particular interest.

• A Master degree in a discipline related to the project is required
• A good track record in research will be evaluated positively.
• Experience in network reconstruction and programming will be considered a plus. 
• Excellent knowledge of English, both spoken and written

6 months