Research Background

This page summarizes selected research backgrounds and outcomes related to SOBIN Institute’s central interests in natural intelligence, decision-making, complex systems, and social optimization.

The topics presented here are not intended as a closed list of research areas. They represent different entry points into a broader question: how hidden computational principles used by nature can be extracted from complex systems and connected to socially meaningful systems.

1. Learning by Physics

1-1) The Tug-Of-War (TOW) principle for efficient decision-making: efficient learning based on physical conservation law and fluctuations

‘Bar Intelligence‘

1-2) ‘The TOW Bombe’ (named after Turing Bombe) for solving societal problems: efficient problem-solving based on physical conservation law and fluctuations

1-3) Decision-making devices based on the TOW principle

1-4) Applications (IoT Communications, Behavioral Economics) based on the TOW principle

  

2. Cryptography and Random Number Generation

2-1) Statistical tests for random number generator (RNG): correction of statistical tests provided by the NIST (National Institute of Standards and Technology, USA)

The N1 Distribution of SHA1 with the corrected threshold and two different theoretical distributions.

2-2) Random Number Generator (RNG)

SEM images of Crystals

  • S. -J. Kim, & K. Umeno: Randomness Evaluation and Hardware Implementation of Non-additive CA-based Stream Cipher, J. Signal Processing 9(1), pp. 71-77 (2005)

2-3) Chaos-based Communication System

3. Random Walk Approach for Time Series Analysis in Finance

3-1) Theory of random walk with chaos

ETMSDs of a random walk using a time series generated by an Aizawa map with B = 1.7 (black circles) and B = 2.2 (red squares). The slope of each ETMSD indicates superdiffusion.