[Pycon] [new paper] "Luca Pappalardo" - Human Mobility Analysis: from the theory to the practice

[info a pycon.it]

Title: Human Mobility Analysis: from the theory to the practice
Duration: 240 (includes Q&A)
Q&A Session: 0
Language: en
Type: Training

Abstract: The availability of human mobility big data (e.g., GPS traces, mobile phone records, social media records) is a trend that will grow in the near future. In particular, this will happen when the shift from traditional vehicles to autonomous, self-driving, vehicles, will transform our society, the economy and the environment. For this reason, understanding the fundamental patterns underlying human mobility is of paramount importance for many present and future applications such as traffic forecasting, urban planning, estimating migratory flows, epidemic modeling, and more.

In this training we will present, with a strong focus on code implementation, an overview on the fundamental principles underlying the analysis of big mobility data. Starting from human mobility data describing the whereabouts of individuals on a territory for a large-enough observation window, we will drive the audience through the extraction of mobility patterns and measures by using a specific Python library designed by the tutorial presenters. This library allows the user to: (1) analyze mobility data by using the main measures characterizing human mobility patterns (e.g., characteristic traveled distance, mobility motifs, predictability of trajectories); (2) simulate individual and collective mobility by executing the most common human mobility models; (3) compare all these models by a set of validation metrics taken from the literature. Since it is supposed to be a practical hand-on tutorial, for every model presented during the training we will show a practical code example presented through the Jupyter notebook.

In detail, the training will cover the following aspects:

 - **The human mobility data landscape**
A natural starting point is to describe the nature of empirical data which has been used in human mobility analysis. We outline the main sources available for human mobility and the relevant information that can be extracted from them. Practical examples will be provided to show how to read and clean mobility data.

 - **Under the microscope: Extracting individual and collective mobility patterns**
We will present some of the fundamental measures and representations used to characterize human mobility both at individual and collective level, such as trip distance, characteristic distance, predictability of trajectories, origin-destination matrix, mobility motifs, and more. With practical code examples we  will show how to extract these patterns from mobility data..

 - **Agents on the move: generating synthetic human mobility data**
Human mobility big data is generally owned by private companies, meaning that they are rarely publicly available. This part will show how to use state-of-the-art models to generate synthetic mobility data, i.e., mobility data representing trajectories which are realistic in reproducing key statistical patterns of human mobility. All the models that will be introduced in this section will be then used in some code examples thank to the mobility library developed by the presenters of this tutorial. 

No specific knowledge about human mobility is required. Basic understanding of Pandas and Numpy are recommended.

Tags: [u'analytics', u'statistics', u'scikits', u'mathematical-modelling', u'bigdata', u'simulation', u'scientific-computing']