Computational IntelligenceUniversity of Oldenburg
I develop computational intelligence methods that connect optimization, learning, cognition, and biological data.
My work investigates how adaptive systems search, learn, reason, and extract structure from complex data.
HOVER A NODE TO INSPECT A DIVISION
01 / Research Divisions
The lab studies intelligent systems across optimization, learning, cognition, and biological discovery. These divisions are distinct, but designed to exchange methods and ideas.
01
Developing robust methods for black-box optimization, self-adaptation, and search in complex, high-dimensional spaces.
02
Designing learning systems that discover useful representations, generalize reliably, and remain interpretable under real-world conditions.
03
Studying computational models of reasoning, learning, decision-making, and the relationship between artificial and human intelligence.
04
Applying machine intelligence to biological data to uncover patterns, generate hypotheses, and support scientific discovery.
02 / Selected Systems
01 / ADVERSARIAL CRITIQUE · LLM PIPELINE · STATISTICS AUDIT
An adversarial AI critic that stress-tests research papers—surfacing overclaims, questionable statistics, methodological weaknesses, and buzzword-heavy writing.
PEER REVIEW, WITH THE POLITENESS FILTER REMOVED.
02 / PARALLEL TRANSLATION · IDIOM NOTES · 99 LANGUAGES
A world dashboard for translation: one phrase rendered simultaneously across the eight most spoken languages, with notes on idioms and grammar — plus a Europe view and a 99-language mode.
ONE PHRASE, 99 LANGUAGES, SIDE BY SIDE.
03 / NUMPY FROM SCRATCH · (1+1)-ES · SVM · MLP · CNN
A hands-on course that implements ML algorithms from scratch in NumPy — SVM, MLP, clustering, and CNNs — and trains every model with the same derivative-free optimizer: the (1+1)-Evolution Strategy.
EVERY MODEL, ONE OPTIMIZER, NO BLACK BOXES.
03 / Publications
04 / Books
Books by Oliver Kramer, published with Springer — from self-adaptive heuristics and evolution strategies to machine learning and the essentials of artificial intelligence.
FIG. B12026
Springer
Springer Link
FIG. B22017
Studies in Computational Intelligence 679
Springer Link
FIG. B32016
Studies in Big Data 20
Springer Link
FIG. B42014
SpringerBriefs in Applied Sciences and Technology
Springer Link
FIG. B52013
Intelligent Systems Reference Library 51
Springer Link
FIG. B62009
Informatik im Fokus
Springer Link
FIG. B72008
Studies in Computational Intelligence 147
Springer Link
05 / Writing
Long-form articles by Oliver Kramer published in Towards Data Science — explaining cognitive prompting, evolutionary optimization, dimensionality reduction, and deep learning for a broad technical audience.
FIG. W12024
How structured, human-like cognitive operations — goal clarification, decomposition, filtering, integration — can be turned into prompts that make language models reason more systematically.
Towards Data Science
FIG. W22023
A selection scheme that spreads a population evenly along the Pareto front using rake lines, keeping multi-objective search diverse without expensive density estimation.
Towards Data Science
FIG. W32023
Turning regression models inside out: latent coordinates are optimized instead of fitted, yielding a flexible framework for manifold learning and embeddings.
Towards Data Science
FIG. W42022
A guided path through the architectures that built modern vision — from the perceptron and LeNet to ResNet and DenseNet — and the ideas that made each step work.
Towards Data Science
06 / Art & Music
Experiments in AI-generated imagery and sound released under the alias ACIDBOT — the same generative machinery as the research, tuned for aesthetics instead of benchmarks.
FIG. A1 — ACIDBOTVIDEO
AI-generated cover art · music video
An ouroboros of fire as cover art for a generative track: the loop that consumes and regenerates itself — the same feedback motif that drives an evolution strategy, rendered in sound and image.
Watch on YouTube
FIG. A2 — ACIDBOTVIDEO
AI-generated music video
A saturated post-urban landscape overgrown by branching structures — diffusion models pushed until the colour space itself becomes the subject. Visuals sequenced to an electronic score.
Watch on YouTube07 / About
Oliver Kramer is Professor of Computational Intelligence at University of Oldenburg, working at the intersection of evolutionary computation, machine learning, cognition, and bioinformatics.
He develops computational methods for understanding and building adaptive intelligent systems — from self-adaptive search in high-dimensional spaces to interpretable learning models and computational accounts of reasoning. His group teaches and supervises across the Department of Computing Science, and collaborates with partners in the natural and life sciences.
08 / Press
Selected media coverage and public conversations on artificial intelligence, cognition, and the future of intelligent machines.
FIG. P1 — DER SPIEGELPodcast »Moreno+1«
KI-Professor Oliver Kramer beschäftigt sich mit Algorithmen, »die menschenähnliche kognitive Leistungen vollbringen«. Für ihn steht fest: Maschinen werden intelligenter als Menschen sein. Wir sollten darüber nachdenken.Read on DER SPIEGEL
09 / Contact
I welcome research collaborations, interdisciplinary projects, doctoral inquiries, and conversations about computational intelligence.