This is a landing page for the workshop “Hands-on machine learning in ecology”.
If you are visiting the webpage but you have not registered yet, unfortunatelly the workshop is now at full capacity. However, given popular demand, we have decided to open the last day of the workshop to all, since it will consist of a set of talks by researchers showcasing their research and applications. We will advertize the Friday Zoom link here (and only here), but to decrease the chance of getting undesired trolling presences, the link will only be available here a few minutes before the Friday event starts.
Open Day Friday 10th 2021 Zoom link now available:
Password: 412579
This workshop is a joint effort by the Centre of Statistics and its Applications (CEAUL), Portuguese Institute for the Sea and the Atmosphere (IPMA), Centre for Research into Ecological and Environmental Modelling (CREEM) at the University of St Andrews and School for Data Science and Computational Thinking at Stellenbosch University, with the kind contribution of all the colleagues below with theirs talks.
- Monday to Friday, 6th to 10th of September 2021, 9 am to 12 am
How it started:
This event started as a desire from CEAUL to provide machine learning training to its members, and has slowly but steadily evolved into a workshop coined “Hands-on machine learning in ecology”. It is a free event,organized an run by kind folks donating their time to do so. We invited Carl Donovan (CREEM), Ian Durbach (CREEM) and Emmanuel Dufourq (Stellenbosch University) to deliver the hands-on component, and then invited a number of researchers to provide talks on applications on ML. This mix of lectures, hands-on applications and talks will hopefully provide a good learning environment, generating a good balance between learning and discussion. We hope you will enjoy it.
How it evolved and the open day on Friday:
Given the interest that was generated around the event, we decided to make it an hybrid event, where we opened the last day (Friday 10th September 2021) to all interested in the topic. On that day the format is essentially expositive, with a number of talks and discussions on the topic. This day is hoped to be somewhat self-contained as a mini symposium with talks on the topic, so could act as a standalone. Nonetheless, bear in mind, While the talks could be seen as research talks, the entire event is a 5 day event and some things on the open day might make sense only within the full scope of the workshop.
Details about the workshop:
Workshop Requirements: Previous knowledge in R and linear models is required; you need to have R and R studio installed for the hands on component.
Level: Introductory
Organization: Marta Rufino (IPMA; CEAUL) & Tiago Marques (CEAUL;CREEM)
Teachers:
Ian Durbach (Centre for Research into Ecological and Environmental Modelling at the University of St Andrews; CREEM)
Carl Donovan (Centre for Research into Ecological and Environmental Modelling at the University of St Andrews; CREEM)
Emmanuel Dufourq (Stellenbosch University, department of industrial engineering & school for data science and computational thinking; African Institute for Mathematical Sciences, junior research chair in data science for climate resilience)
Program:
Monday:
Monday, 6/9/2021 9:00-9:15 h
Introduction of the teachers and workshop (Tiago Marques & Marta Rufino)
Monday, 6/9/2021 9:15-10:00h
Gciniwe Dlamini (IBM Research Africa; 30 min talk + 5 questions + 5 break)
Title: Applications of Machine learning in Healthcare
Monday, 6/9/2021 10:00-10:30h
Ian Durbach previous work (20 min talk + 5 min questions + 5 min break)
BREAK 15 min
Monday, 6/9/2021 10:45-11:15h
Carl Donovan previous work (20 min talk + 5 min questions + 5 min break)
Monday, 6/9/2021 11:15-11:45h
Emmanuel Dufourq previous work (20 min talk + 5 min questions + 5 min break)
WRAP UP 15 min
Tuesday (Carl Donovan)
introduction to machine learning
trees (CART-style, bagging, boosted regression trees, random forests)
Wednesday (Ian Durbach):
feedforward neural networks
convolutional neural networks
Thursday (Emmanuel Dufourq):
transfer learning
implementing deep learning models (choosing architectures, hyperparameters)
Friday (talks on applications: 20 min + 5 min talks + 5 min interval = 30 min blocks) – Open Day!
Friday, 10/9/2021 9:00-9:30h
– Ketil Malde (Institute of Marine Research, Bergen, Norway; Department of Informatics, University of Bergen, Norway; https://www.hi.no/hi/om-oss/ansatte/ketil-malde)
Title: Artificial intelligence and the future of marine resource management;
Friday, 10/9/2021 9:30-10:00h
– David Harris-Birtill (https://sachi.cs.st-andrews.ac.uk/people/faculty/david-harris-birtill/ ; https://medtech.cs.st-andrews.ac.uk; Director of Digital Health MSc and Lecturer in Computer Science at the University of St Andrews);
Title: Deep Learning for Cancer Detection in Medical Imaging;
Friday, 10/9/2021 10:00-10:30h
– Benito Blas (Department of Ecology & Multidisciplinary Institute for Environment Studies “Ramon Margalef”, University of Alicante; https://www.blasbenito.com/);
Title: Incorporating autocorrelation in spatial models produced with ML
BREAK 10 min
Friday, 10/9/2021 10:40-11:10h
– Bruno Caneco (Marine Scotland Science – Freshwater Fisheries Laboratory);
Title: Salmon Scale Imaging: Automating the extraction of growth information using a deep-learning object detection algorithm);
Friday, 10/9/2021 11:10-11:40h
– Rafaela Cruz (Faculdade de Ciências e Tecnologia, Universidade NOVA de Lisboa);
Title: Machine Learning Models Applied to Harmful Algal Blooms and Shellfish Contamination Forecasting
Friday, 10/9/2021 11:40-12:00h
CONCLUDING REMARKS