Email : reza.khorshidiATgeorgeinstitute.ox.ac.uk
Experienced in building and leading teams of elites for helping businesses devise the right big-data machine learning strategies; delivered multiple algorithmic software products and consulting across Healthcare/Biotech, Finance/Insurance, Retail and Telco.
Team building, leadership, management consulting, SaaS, PaaS, big data analytics (e.g., Hadoop and Spark), machine learning (e.g., Bayesian inference, Gaussian processes, regression, classification, clustering, hierarchical modelling), biomedical informatics, signal and image processing, graphical/network modelling, programing (e.g., Python, R, C++, SAS, Matlab, Linux, SQL).
Head of Quantitative Analytics, EMEA and Special Project Lab | AIG (Nov.12-Present)
I built, manage, train and lead teams of data elites (including quants and data scientists) to deliver disruptive solutions that help AIG (the biggest property and casualty insurer in the world) devise the right big-data strategy and infrastructure. My work consists of leading several projects concurrently, understanding and advising individual business units, and rapid implementation of software solutions that have machine learning at their core. I also lead the Special Projects Lab; a transformational research unit, carrying out research on deep learning, computer vision, natural language processing, and signal processing (e.g., speech and tone analyses); these are disruptive solutions that the insurance industry is seeing for the first time.
Program Lead, Machine Learning and Biomedical Informatics | The University of Oxford (Oct.12-Present)
I lead the machine-learning program at The George Institute for Global Health aimed at utilising big data for biomedical informatics research, improving healthcare delivery, and disease management. This involves developing coherent research projects, supervising postgraduate students and post-docs, and leading the statistical analyses of largest healthcare and bioinformatics datasets. I also collaborate with the analysis group of FMRIB centre and am part of the human connectome project (an ambitious international consortium focused on mapping the structural and functional networks in the brain).
Analytics Lead | Opera Solutions (Oct.10-Oct.12)
I managed the development of machine learning solutions for consulting clients in government, healthcare, telecommunication, retail and media with innovative, powerful, and profit-enhancing solutions that had big-data in their core. Examples include state-of-the-art recommendation algorithms and relevance engines, various time series models, variety of probabilistic risk models, and wide range of segmentation/clustering/classification and missing-data solutions.
Computational Neuroimaging Researcher | GlaxoSmithKline & The University of Oxford (Oct.07-Oct.12)
Funded by GSK and based in in The University of Oxford, I developed statistical models and software solutions for explaining correlation structures in the brain’s spatiotemporal data, mining large-scale neuroinformatic databases, and separating signal from noise. The main keywords of my research were functional connectomics, Gaussian processes, Bayesian inference, graphical modeling / causal inference, classification, regression, pattern recognition, modeling nonstationarity, and multivariate pattern analysis.
Project Manager / R & D Engineer | Tamin Mgmt. Consultancy & Computing Services (May.04-Apr.07)
I managed teams of healthcare, insurance, software and networking professionals in order to provide the largest network of healthcare centers in Iran with state-of-the-art software, networking, and automation solutions.
The University of Oxford, UK (Oct.07-Oct.10): DPhil in Computational Neuroscience & Statistics (funded by GSK for £150K)
Amirkabir University of Technology, Iran (Oct.99-Oct.06): M.Sc. in Biomedical Eng. & B.Sc. in Electrical Engineering
Beyond (London, UK; Jun.11-Sep.11): As the ‘head of quantitative strategy’, I used R for developing sophisticated models and analytics procedures for mining social media data and survey (large-scale categorical data) analysis.
Oxford Internet Inst. (Oxford, UK; May.11-Sep.11): Modeling the citation universe of the US Supreme Court, using social network analysis techniques (in R & Python), in order to mine different clusters of ideologies, and their evolution over the past 150 years.
RIKEN Brain Science Inst. (Saitama, Japan; Mar.07-Oct.07): Modeling multi-layer cortical neural networks.
Inst. Of Theoretical Physics & Mathematics (Tehran, Iran; Feb.06-Apr.07): On an artificial Intelligence internship, working on machine vision algorithms that are biologically plausible.
Iranian Ministry of Health (Tehran, Iran; May.03-Mar.07): Investigated biomedical patents for IP/novelty.
Tech/Biotech Startups, World News/Politics, Traveling, Food, Sports, Music and Cinema
(1) Automatic Denoising of Functional MRI Data: Combining Independent Component Analysis and Hierarchical Fusion of Classifiers, NeuroImage 2014
(2) Using Gaussian-Process Regression for Meta-Analytic Neuroimaging Inference Based on Sparse Observations, IEEE Trans. On Med. Imaging 2011
(3) Adjusting the effect of nonstationarity in cluster-based and TFCE inference, NeuroImage 2011
(4) Network modelling methods for FMRI, NeuroImage 2011.