Guozhong (Gordon) An Email & Phone Number

Amersfoort, Nl

Within the Stater Datalab, focusing on developing leading edge machine learning and AI models with diverse use cases involving deep learning and NLP, predictive modelling while remaining active in risk modelling.

Amersfoort, Utrecht, Netherlands

• Responsible for full life cycle of predictive models from model conception and data preparation to model development, validation, and implementation as well as annual reviews, updating, and monitoring• Developing models for leading mortgage lenders (banks and insurance companies) in the Netherlands. Models include credit risk (PD and LGD), credit risk scoring, prepayment, fraud risk, and mortgage pipeline risk and arrears collection. • Exploring and applying machine learning techniques such as sentiment analysis and topic modelling (LDA) to customer service communications • Coaching and supervising junior team members and M.Sc. thesis students of Econometrics or Statistics

Dubin, Ireland And Gent Belgium

Researched and developed several leading-edge machine learning algorithms including: regularized linear regression, SVM, decision trees, k-means clustering, association rules and backpropagation neural networks. These algorithms form the key intelligence in Norkom's flagship product called Alchemist. Norkom's Software is used by financial and telecom companies for data miningLed predictive modelling and consulting projects for international clients including Citibank, Fortis Bank, Bank Leumi, and Sopres with measurable ROI exceeding client expectation. Led a team of highly qualified (Ph.D. level) modelling analysts.Built up a strong modelling team in a start-up environment, while playing hands-on role in model development and project delivery to clients including: Actel (increased direct mailing response rate by 56% overall), Neckermann, Overtoom (Customer churn modelling) and SMAP (real-time scoring model for candidate selection and scripting in telemarketing).

Rijswijk, Netherlands

Conducted research in machine learning, neural networks, and genetic algorithms. My publication on the effect of adding noise to neural networks is still being referenced in today’s scientific research.Developed and prototyped several pattern recognition applications of neural networks including the automatic recognition of events from seismic signals. Pioneered research of convolutional neural networks applied to the inversion of well log signals. Convolution neural networks form the basis in present time deep learning applications.Managed extramural neural network research projects at Leiden University and Radboud University Nijmegen and supervised several M.Sc. thesis projects of students.

Leiden, The Netherlands

Pioneered quantum fixed node Monte Carlo simulation on lattice model. Published a number of papers on two-dimensional Hubbard model. Collaborated with Prof. Hans Van Leeuwen in this period.

Ph D, Physics

Bsc, Physics