Michael Nielsen
Specialkonsulent at Skattestyrelsen
San Francisco, California, Denmark
Summary
🤩
Rockstar🎓
Top SchoolMichael Nielsen is a Danish data-driven special consultant with 17 years' experience optimizing research and tax collection processes across public and private sectors. Currently a process owner at the Danish Tax Agency in San Francisco, he blends economics training from Copenhagen Business School with hands-on data consulting to improve compliance, timeliness and operational efficiency. Previously he led market analysis programs and complex multi-year projects at Kantar, delivering actionable insights and streamlining reporting pipelines. A seasoned questionnaire programmer and dashboard implementer from his Catinét days, he pairs technical fluency with stakeholder-facing delivery. Michael also contributes to open-source ML education—maintaining code samples for a well-known neural networks textbook—revealing a curiosity-driven approach summarized on GitHub as "searching for the numinous." Colleagues would describe him as a pragmatic problem-solver who brings analytical rigor and a taste for foundational ML concepts to process design.
17 years of coding experience
3 years of employment as a software developer
Master, Economics, Master, Economics at Copenhagen Business School
Danish, English
Stackoverflow
1reputation
0reached
0answers
0questions
Github Skills (10)
neural-network10
machine-learning10
python10
data-science10
backpropagation10
numpy10
matplotlib10
scikit-learn9
scikit9
theano8
Programming languages (5)
Github contributions (5)
mnielsen/neural-networks-and-deep-learning
Aug 2012 - Mar 2018
Code samples for my book "Neural Networks and Deep Learning"
Role in this project:
ML Engineer & Data Scientist
ML Engineer & Data ScientistContributions:225 commits, 25 PRs, 20 pushes in 5 years 7 months
Contributions summary:Michael primarily worked on code samples for a neural network and deep learning book. Their contributions focused on implementing fundamental concepts, including perceptron learning, sigmoid and step functions, and basic plotting in Python using Matplotlib. They also developed code for an XOR function using gradient descent and started work on an MNIST digit recognition system using backpropagation, including comparison baselines like average darkness and SVMs.
deep-learningneural-network
mnielsen/notes-on-writing
Jan 2013 - Jun 2020
Notes on writing
Contributions:63 commits, 7 PRs, 3 pushes in 7 years 6 months
notes
Find and Hire Top DevelopersWe’ve analyzed the programming source code of over 60 million software developers on GitHub and scored them by 50,000 skills. Sign-up on Prog,AI to search for software developers.