## Learning Machines, Extended Logic, & Intelligence

This page contains loosly connected research lines on machine learning, information theory, as well as artificial and other intelligence. Learning Machines better reason according to logic. If uncertainties are involved, this should be extended, probabilistic, or Bayesian logic. The same is true for any form of intelligence, whether of human, artificial, or other nature.

## Highlights

**Artificial intelligence combined**

Artificial intelligence expands into all areas of the daily life, including research. Neural networks learn to solve complex tasks by training them on the basis of enormous amounts of examples. Researchers at the Max Planck Institute for Astrophysics in Garching have now succeeded in combining several networks, each one specializing in a different task, to jointly solve tasks using Bayesian logic in areas none was originally trained on. This enables the recycling of expensively trained networks and is an important step towards universally deductive artificial intelligence.

- Learn more

**The embarrassment of false predictions - **

How to best communicate probabilities?

How to best communicate probabilities?

Complex predictions such as election forecasts or the weather reports often have to be simplified before communication. But how should one best simplify these predictions without facing embarrassment? In astronomical data analysis, researchers are also confronted with the problem of simplifying probabilities. Two researchers at the Max Planck Institute for Astrophysics now show that there is only one mathematically correct way to measure how embarrassing a simplified prediction can be. According to this, the recipient of a prediction should be deprived of the smallest possible amount of information.

## Learning Machines

**Geometric Variational Inference**

Philipp Frank, Reimar Leike, Torsten A. Enßlin, submitted arXiv:2105.10470**Bayesian Reasoning with Deep-Learned Knowledge**

Jakob Knollmüller, Torsten A. Enßlin, 2021, Entropy 23, 693 arXiv:2001.11031**Bayesian decomposition of the Galactic multi-frequency sky using probabilistic autoencoders**

Sara Milosevic, Philipp Frank, Reimar H. Leike, Ancla Müller, Torsten A. Enßlin, A&A, 650, A100 arXiv:2009.06608-
**Information Field Theory: from astrophysical imaging to artificial intelligence**

Talk by Torsten A. Enßlin at Joint Astronomical Colloquium (ESO Garching, 13.2.2020) **Sharpening up Galactic all-sky maps with complementary data. A machine learning approach**

Ancla Müller, Moritz Hackstein, Maksim Greiner, Philipp Frank, Dominik J. Bomans, Ralf-Jürgen Dettmar, Torsten A. Enßlin, (2018) Astronomy & Astrophysics, Volume 620, id.A64, 20 pp. doi:10.1051/0004-6361/201833604**Metric Gaussian Variational Inference**

Jakob Knollmüller, Torsten A. Enßlin, submitted arXiv:1901.11033**SOMBI: Bayesian identification of parameter relations in unstructured cosmological data**

Philipp Frank, Jens Jasche, Torsten A. Enßlin, (2016) Astronomy & Astrophysics, Volume 595, id.A75, 18 pp. doi:10.1051/0004-6361/201628393 arXiv:1602.08497

## Extended Logik

**A Bayesian Model for Bivariate Causal Inference**

Maximilian Kurthen, Torsten A. Enßlin, Entropy 2020, 22, 46; doi:10.3390/e22010046 arXiv:1812.09895**Encoding prior knowledge in the structure of the likelihood**

Jakob Knollmüller, Torsten A. Enßlin, submitted arXiv:1812.04403**Towards information optimal simulation of partial differential equations**

Reimar H. Leike, Torsten A. Enßlin, Physical Review E, Volume 97, Issue 3, id.033314 ePrint, arXiv:1709.02859**Optimal Belief Approximation**

Reimar Leike, Torsten A. Enßlin, Entropy 2017, 19, 402; doi:10.3390/e19080402 arxiv:1610.09018**Operator Calculus for Information Field Theory**

Reimar Leike, Torsten A. Enßlin, Physical Review E, Volume 94, Issue 5, id.053306 (2016) arxiv:1605.00660**Inference with minimal Gibbs free energy in information field theory**

Torsten A. Enßlin, Cornelius Weig 2010, Physical Review E 82, 051112 arXiv:1004.2868 and Comment on Paper and Reply to Comment

## Artificial, Human, & Other Intelligence

**Theoretical Modeling of Communication Dynamics**

Torsten A. Enßlin, Viktoria Kainz, & Céline Bœhm, submitted (2021) arxiv:2106.05414

**The Rationality of Irrationality in the Monty Hall Problem**

Torsten A. Enßlin & Margret Westerkamp, Annalen der Physik, vol. 531, issue 3, p. 1800128 (2019) doi: 10.1002/andp.201800128 arxiv:1804.04948