1. Faculty of Social Sciences
  2. People
  3. Profile

Andrew Smith

  • Position: Senior Research Fellow
  • Areas of expertise: Numerical methods, simulation methods, statistical analysis, machine learning, deep learning. Software design and architecture, Research Software Engineering.
  • Email: A.P.Smith@leeds.ac.uk
  • Website: github | LinkedIn | ORCID

Profile

I am a highly experienced quantitative professional with a diverse career spanning computational physics, finance, academic research, and technology.

After completing a Ph.D. in computational physics, I began my career developing hydrodynamic simulations for the UK Government. I went on to spend the majority of my career as a quantitative analyst at various international financial institutions. More recently, I've leveraged my expertise across academic research roles, quantitative and data science positions within tech startups, and independent consultancy.

My academic work has primarily focussed on demographic forecasting and crime modelling, developing and employing advanced techniques such as microsynthesis and microsimulation.

A cornerstone of my profile is my extensive experience in professional software architecture, engineering and development. which I apply extensively to projects in a research setting. I am a prolific contributor to open-source initiatives.

Responsibilities

  • Researching methodologies for quantifying crime concentration and predictivity
  • Data Science Technical Leadership
  • Research Software Engineering

Research interests

A recurrent theme in my career is cross-fertilisation: applying techniques learned and developed in one field to enrich another.

I'm currently working on the Safer Streets project and the work is being documented here: https://safer-streets.github.io

Authorship of a number of open-source software packages is a prominent focus of my academic work, for example:

neworder: a high-performance dynamic microsimulation framework for python. See https://neworder.readthedocs.io
humanleague: a microsynthesis package for python and R, featuring IPF and quasirandom sampling methods. See
https://github.com/virgesmith/humanleague

I've blogged about topics related to Research Software Engineering with a focus on reproducibility, see
https://virgesmith.github.io

My previous work in the School of Geography/LIDA is documented here:
https://www.cdrc.ac.uk/research/population-housing-and-infrastructure/mistral/

Qualifications

  • Certificate in Financial Mathematics, Carnegie-Mellon University
  • Ph.D. Computational Physics, University of Salford
  • M.Sc. Physics, University of Salford
  • B.Sc. Physics, University of Manchester