Experience

 
 
 
 
 
July 2023 – Present
Bergen

Principal developer

Equinor

Developer & Research

  • Researching new algorithms in ensemble based data assimilation
  • The Ensemble Information Filter (EnIF)
  • Cooperation with TU Delft on non-linear and non-Gaussian ensemble methods
 
 
 
 
 
September 2021 – July 2023
Bergen

Senior Consultant

Sonat

Equinor Scientific Computing Team

  • Refactoring data-assimilation algorithms for reservoir engineering.
  • Developed the GraphSPME sparse precision estimation library.
  • Developed the Ensemble Information Filter and Smoother algorithms.

Norwegian Hull Club

  • Workshop on automatic and adaptive gradient tree boosting
 
 
 
 
 
January 2021 – January 2023
Bergen

Adjunct Associate Professor

University of Bergen, Department of Mathematics

  • Research: Developing information theory for automatic ML-algorithms.
  • Teaching and seminars: Course development for actuarial and data-science courses.
  • Master student supervision.
 
 
 
 
 
September 2020 – August 2021
Bergen

Data Scientist & Actuary

Frende

  • Translate business needs into mathematical (optimization) problems.
  • Advocate version control, code-standards, packaging, CI/CD, containerisation.
  • Introduced advanced regression techniques such as GBM and mixed effects GAM.
 
 
 
 
 
September 2017 – August 2020
Stavanger

PhD Candidate

University of Stavanger

Research areas:

  • Information theoretic gradient tree boosting: aGTBoost
  • Saddlepoint addjusted inversion of characteristic functions
 
 
 
 
 
June 2016 – August 2017
Bergen

Actuary

Tryg

Responsibilities include:

  • Extraction, preprocessing, and analysis of large amounts of data
  • Pricing of products and risks
  • Analysis of customer behaviour
 
 
 
 
 
January 2015 – December 2017
Bergen

Research assistant

University of Bergen

Hosted seminars in the “Kaggle club” at UiB

Teaching assistant:

  • STATLEARN: Statistical Learning
  • STAT220: Stochastic Processes
  • MAT102: Elementary Calculus II
  • STAT101: Elementary Statistics

R

80%

Statistics

70%

Coffee & Code

100%

Recent Posts

In my previous post, I mentioned my friend and previous co-worker at Tryg, Ole Schei, that is wonderfully meticulous, and has for some …
7 min read 0 Comments C++, R, Affine latent variables

Student acceptance Lately I got to play with some very interesting data, relating to student acceptance (and ranking) into applied …
10 min read 0 Comments R, Boosting

I was intending to write a longer blog-post about my favourite dataset of all time: the internet surf-times of a previous co-worker …
5 min read 0 Comments R, C++

We seek to employ the framework of the R package Template Model Builder TMB (Kristensen et al. 2016) which approximately “integrates …
13 min read 0 Comments R

Four years ago, back in 2014, a friend from HKUST, that had started working in the finance industry, asked me about the VIX. He could …
4 min read 0 Comments R

Projects

*
All Boosting Data Science

aGTBoost

Adaptive and automatic gradient boosting computations

spaTMB

Using TMB to build the saddlepoint approximation.

SPI

Saddlepoint adjusted inversion of characteristic functions.

Recent Publications

For certain types of statistical models, the characteristic function (Fourier transform) is available in closed form, whereas the …
Preprint PDF Code Dataset

This thesis considers the problem of likelihood- based parameter estimation for time-homogeneous jump-diffusion processes. The problem …
Preprint PDF Code Dataset Project

Recent Talks

More Talks

Gradient boosting has been highly successful in machine-learning competitions for structured/tabular data since the introduction of …
Oct 9, 2019 12:00 PM — 1:00 PM BigInsight Wednesday Lunch

Gradient boosting has been highly successful in machine-learning competitions for structured/tabular data since the introduction of …
Sep 17, 2019 1:00 PM — 2:00 PM The stochastics seminar

In gradient tree boosting, the functional form of the ensemble repeatedly changes during training. To select a sensible functional …
Jun 26, 2019 5:08 PM — 7:10 PM EcoSta 2019

Contact