¹û¶³Ó°Ôº

Module Description
To ensure healthy environments for both people and nature, it is critical to understand the impact of rapid environmental change on biodiversity. For example, how species are adapting to land use and climate change, and how the degradation of nature impacts people and society. New high-resolution data streams produced from a growing array of sensor technologies such as high-resolution satellite imagery, visual and audio sensor data, geospatial tracking devices and environmental DNA are revolutionising how we are monitoring our natural environment. At the same time, breakthroughs in Artificial Intelligence and data science are transforming society. However, the challenge remains: how do we utilize these cutting-edge tools to analyse this vast ecological and environmental data and shape our understanding of ecosystem management and address global crises such as biodiversity loss and climate change mitigation and adaptation?

Module Aim
This module aims to bridge the gap between computer science and ecology to advance your knowledge and analytical skills to infer wildlife abundance, occupancy and distribution trends from high resolution data sensor streams using AI & other advanced spatiotemporal statistical tools in R and Python.

Learning Objectives
At the end of the module students will be able to:

• Learn different data types & limitations from range of ecological sensors
• Learn key concepts and applications of machine learning to process ecological data and evaluate models
• Learn key concepts and applications in spatial & temporal modelling to infer and visualize biodiversity data & critically assess analyses in relation to experimental design and biases
• Practice and develop your data science and statistical modelling analysis skills through practicals and problem-based assessments in R and Python, write up findings and put them in context
• Practice ethical, secure, and FAIR use principles of data collection and management

Each week the module will cover an aspect of data analysis using examples and exercises drawing on environmental data. Topics will include:

• Introduction to Python Coding Environment
• Introduction to Machine Learning and Deep Learning, including supervised learning (regression and classification), unsupervised learning (dimensionality reduction and clustering) and model evaluation
• Application of machine learning to processing wildlife Images, bioacoustics, remote sensing and biologging movement data
• Introduction to Bayesian spatiotemporal modelling
• Application of Bayesian modelling to wildlife occupancy, abundance across space and time
• Introduction to data visualisation, ethics and FAIR use of data.

There will be weekly office hours where students can attend to ask and get help with any coding/statistical issues they have encountered.