Hello, I'm Jourdain
I'm currently studying an MSc in Remote sensing at the University College London. I previously studied at the University of St Andrews and achieved a BSc (Hons) in Geology.
I am interested in leveraging Machine Learning and Deep Learning to solve problems in Geosciences, and help us to better understand our planet.
Outside of academics, I like to spend most of my time outside, rock climbing, mountaineering and cycling.
Skills : Python | R | Java | Latex | Deep Learning Keras/Tensorflow & Pytorch
Software : ArcGIS Pro | Arcmap | Adobe (Photoshop and Illustrator) | Excel | QGIS | Google Earth Engine
Fellow of the Geological Society of London
LinkedIn | Github | CV | Email: Jourdain@mcilquham.com | Number: +44 7549876875
My Portfolio
Automated Mineral Maps on Mars
Pytorch | Python | Deep Learning | GAN
For my Masters dissertation I developed a Pytorch Model to automate mineral mapping of hyperspectral data on Mars. By implementing a Generative Adversarial Network, 27 minerals could be identified by their spectral features
Cartography Portfolio
Cartographic Principles | ArcGIS | Python
Using ArcGIS and Python I have produced several maps on different Social, Economic and Geospatial topics.
Predicting Snow Cover From Modis Data
Python | Times Series | Web Scraping
Using Python I built a model that requests Modis data based upon the tile id. This data is then processed to calculate the snow cover, to visualise how the melting of the snow will impact a water table.
Discovering the Geology at the recent Lunar Landing sites
R | Python | Hyperspectral | Geology
For my undergraduate dissertation I used ArcMap and Python to create geological maps of the recent Chang'e landing sites on the Moon. This project was supervised by Prof Bernard Foing. Subsequently, I was able to publish an abstract of my work and present at the European Geosciences Union conference.
LSTM model predicting the location and size of Wildfires in California
LSTM | R | ML | ArcGIS
Using an LSTM and k-means clustering a map showing the predicted location of california wildfires was made alongside a predition on their size.
Machine Learning and Spatial Analysis of Review Data
In Python I used Natural language processing to understand venue review data. My aim was to understand the relationship between the text review data and the venue rating. Following clustering of review data several maps were created showing the location of several positive review locations.
Mobility Assessment of Cambridge
R | Python | Social Analysis | Spatial
Using a now defunct social networking application called Gowalla, the aim was to map and understand the mobility patterns of users. The Gowalla dataset allowed users to check-in to locations. Pairing this location based data with Centrality maps created for Cambridge, allowed the mobility patterns of users to be understood. Using this information, the optimal location for a new venue could be determined.
Member of the EuroMoonMars team
Research Group
Duirng my undergraduate dissertation I was a member of the EuroMoonMars team. This was sponsored by the International Lunar Exploration Working group founded by several of the worlds space agencies ( ASA, ASI, BNSC, CNES, DARA, ESA, ISAS, NASA). This experience enabled me to present my disseration at several workshops and a conference. It also enabled me to meet and discuss ideas with fellow students and researchers.
