[Forum SIS] Call for candidates: PhD grant for 3 years, starting between October 2020 and February 2021

[Andrea Gilardi]

 Supervisor: Prof. Jorge Mateu, Department of Mathematics, University Jaume
I of Castellon Spain). Visit http://www3.uji.es/~mateu.

Salary: 1300 euros/month approx

Topic: Statistical analysis of events in space and time with a focus on
networks and trajectories

Research:

New technologies have witnessed the era of massive data (big data) and with
it the emergence of data science field. In a massive world of information,
the need for statistical treatment of events that evolve in space-time over
a network system is increasingly common. When these events are monitored
over time they give information about the movement and provide trajectories
over networks. The theory of space-time stochastic processes on networks
provide theoretical and methodological support for this type of data.

This project aims to address the statistical modeling of this type of
events in these new scenarios. We start with parametric analysis of first
order characteristics, which are necessary to build second order and higher
order ones. We intend to characterise all orders (as far as possible) of a
point process on networks to determine the underlying structure of
space-time interaction existing on the network. With this, we propose
parametric theoretical models of specific processes on networks beyond the
Poisson models. One of these models is Cox processes on networks that
combine random fields (geostatistics) with point processes. We will develop
Bayesian inference methods for these models. These constructions open the
way to the analysis of trajectories. We will first determine simple
trajectories of origin-destination type by proposing Bayesian probabilistic
models. We will extend this context to the case of marked functional
processes where the marks will represent the trajectories.

The project is motivated by real problems from the field of Criminology,
where the events are types of crimes occurring in cities, traffic
accidents, movement of people within a city, or car thefts and their
recoveries. We will provide probabilistic predictive models for the
criminal activity over city networks. We will build our own software that
can be used by external institutions (EPOs) that support the project.


Requirements for candidates:

1) Undergraduate in Mathematics and/or Statistics (main option). Computer
Science could also be accepted as a secondary option.
2) Master in Maths, statistics, computer Science or related disciplines.
3) The undergraduate degree has to be obtained from 2016 to now.
4) Knowledge on stochastic processes, spatial and spatio-temporal
statistics, and R-coding will be appreciated  and up-weighted in the
candidate assessment.

Deadlines: Official call will be open around 23-24 June and it will remain
open for only the next 10 days.

Contact: Those candidates interested in applying, please send an email to
mateu a uji.es as soon as possible to fix details of the application.
-------------- parte successiva --------------
Un allegato HTML è stato rimosso...
URL: <http://www.stat.unipg.it/pipermail/sis/attachments/20200620/6be198f4/attachment.html>