Kaisa_2012_3_photo by Veikko Somerpuro

Enrol
4.2.2020 at 11:00 - 6.3.2020 at 15:00

Timetable

Here is the course’s teaching schedule. Check the description for possible other schedules.

DateTimeLocation
Wed 1.4.2020
09:30 - 12:30
Wed 1.4.2020
14:00 - 17:00
Thu 2.4.2020
09:30 - 12:30
Thu 2.4.2020
14:00 - 17:00
Fri 3.4.2020
09:30 - 12:30
Fri 3.4.2020
14:00 - 17:00

Description

PLEASE NOTE THAT THIS COURSE WILL BE HELD COMPLETELY ONLINE!

Doctoral candidates

Networks offer a convenient way to represent and investigate countless real world systems where entities of some kind (i.e. network nodes) are connected by a static relationship or a dynamic process (i.e. network links). As it is becoming clearer that the dense – yet often elusive - networks of ecological interactions and dependencies permeating complex natural systems play fundamental roles in the emergence and maintenance of biodiversity, there is no surprise in the increasing popularity of network analysis as a tool to investigate fundamental issues in ecology and biological conservation. Much attention has been directed towards the investigation of networks depicting mutualistic and antagonistic interactions. In particular, several studies have tried to understand how - and to what extent – specific, non random structural patterns observed in real world networks contribute to the robustness (or determine the fragility) of natural systems. Notably, most of the knlowedge and tools permitting to tackle those (and additional questions) have been developed in scientific contexts other than ecology, and particolarly in the fields of physics and mathematics. Although this offers a great opportunity for interdisciplinary, it also generates risks due to the possible misuse of techniques (and/or misinterpretation of results) when users (ecologists) try to skip bases and jump into network analysis without first building solid theoretical bases.

This course aims at introducing participant PhD students to ecological network analysis while minimizing this risk, thanks to the synergic effort of a network physicist (C.C.) and a macro-ecologist (G.S.). Ideally, the course will provide participants with basic – yet formally accurate - knowledge of fundamental network science concepts, as well as with practical tools and guidelines to apply those concepts to the investigation of specific natural systems. Theoretical lessons will be paired with practical exercises and coding training sessions (in R). At the end of the course, we expect students to be able to design and conduct independendently basic network analyses to tackle specific ecological questions relevant to their own research lines. More than this, our ambition is that of providing them with enough background information and tools to start individual learning paths in ecological network analysis in a successful and productive way.

To be completed at any time during doctoral studies

Course structure/topics

1. Brief introduction to network theory

background history
basic definitions
networks in the real world (ecology and beyond)

2. Studying networks

formal network representations (e.g., edge list versus adjacency matrix)
basic network features (e.g., edge directionality and weight, unimode versus bipartite networks)
basic local and global network structural properties (node degree – degree distribution; path length – diameter; clustering; k-coreness; edge betweenness; PageRank…)
different kinds of networks (e.g. regular, random, power law, small world)

3. Networks in ecology: not just food webs and ecological networks

history of food-web ecology
mutualistic networks and biodiversity
investigating network structure to compare natural systems
robustness and fragility of networks to loss of species and/or interactions

4. Network epidemiology

using networks to map dynamic infectious processes
basic epidemiological models and their implementation in networks
examples of application

5. Network visualization

introduction to network layout
beyond the hair-ball: challenges in visualizing large networks
layout optimization algorithms
customized layouts
three-dimensional visualization and animation

As minimum expectation, attendants will learn to:

download ecological network data from various online datasets
load those data in proper network format
generate random network with given properties
measure local and global network properties (and rank nodes according to different criteria of importance)
manipulate/modify networks (add and remove nodes and edges; extract sub-networks; randomise links)
evaluate network robustness/simulate network collapse following the loss of nodes
visualise networks (also using animation techniques)

Most of the training will be done in the R programming language. Basic knowledge of R and R Studio is therefore desirable. Other software tools will be introduced as well.

As agreed with the supervisor or person in charge of studies

As agreed with the supervisor or person in charge of studies

Full attendance required

Language: English or as agreed with the supervisor or person in charge of studies

As agreed with the supervisor or person in charge of studies

Teachers: Giovanni Strona (REC, University of Helsinki) and Claudio Castellano (Italian National Research Council, Rome)

Contacts:

Giovanni Strona (giovanni.strona@helsinki.fi)

Claudio Castellano (claudio.castellano@roma1.infn.it)