Adam Hornsby

data & cognitive science

I'm a part-time Experimental Psychology PhD Student @ University College London and a Senior Data Scientist @ dunnhumby.

I'm interested in human decision-making and the ways it can be improved. My PhD research (in collaboration with Professor Brad Love and dunnhumby) aims to understand how consumers develop preferences for products in the supermarket. I combine experimental psychology, big data analyses and cognitive modelling.

At dunnhumby, I develop and deploy machine learning models at scale. I'm particularly interested in model deployment, natural language processing and reinforcement learning algorithms.

If you would like to get in touch, please email me.


Several in preparation.

[ Link ] Hornsby, A.N. & Love, B.C. (2014). Improved Classification of Mammograms Following Idealized Training. Journal of Applied Research in Memory and Cognition. 3, 2, 72-76.

Keynotes and blogs

Understanding Customers Better Through Neural Network Embeddings @ DataMashup, London. (21/06/18)

[ Blog ] Solving the Problem of Product Similarity with Data Science, (03/05/18)

[ YouTube, Abstract ] Understanding Customers Better Through Neural Network Embeddings @ Data Science Festival, London. (21/04/18)

[ Abstract ] Understanding Customers Better Through Neural Network Embeddings @ RE.WORK Deep Learning in Retail Summit, London. (15/03/18)

What is Customer Data Science? @ University of Edinburgh, CDT Data Science Day. (14/10/15)


[ Article ] DataIQ, New Talent Award (2018)

[ Profile ] Kaggle, Competitions Master (2016)

UCL, WHR Rivers Prize (i.e. 1st place undergraduate in Psychology) (2013)

UCL, Dean's List (2013)

Taught courses

Applied Machine Learning @ dunnhumby

Work & education

Senior Data Scientist @ dunnhumby. (2014-)

Experimental Psychology, PhD (part-time) @ UCL (supervised by Professor Brad Love). (2016-)

Psychology, BSc (Hons) @ UCL, 1st (dissertation supervised by Professor Brad Love). (2010-2013)

Side projects

[ Code ] MNIST pixel in-painting (experimenting with RNNs in TensorFlow). ~2017

[ Viz, Code ] Visualisations of taxi journeys in Porto (experimenting with d3). ~2015