• Home
  • Iaroslav Omelianenko
Iaroslav Omelianenko

Iaroslav Omelianenko
NewGround · Research and Development

Master of Engineering

About

31
Publications
4,629
Reads
How we measure 'reads'
A 'read' is counted each time someone views a publication summary (such as the title, abstract, and list of authors), clicks on a figure, or views or downloads the full-text. Learn more
3
Citations
Introduction
Iaroslav Omelianenko is a CTO and Research Director at the NewGround LLC. His research interests include human-computer interaction, genetic algorithms & neuroevolution, synthetic intelligence, reinforcement learning, control & optimization, neurobiology and natural evolution. He leads the Research and Development team, which applies genetic algorithms to implement control strategies with a minimal computational footprint that can solve a variety of control & optimization tasks as well as do research in brain-computer interfaces. He has more than 30 years of experience with software design and project management. Actively participates in open source projects. He presented research papers as an author at international conferences.
Additional affiliations
July 2006 - January 2022
NewGround LLC
Position
  • Research Director
Description
  • As a research director, I supervise research department and actively participate in research activities. My main area of interest is genetic algorithms, neuroevolution, and swarm intelligence.
Education
September 1994 - June 1999
Ukrainian State Universtity of Food Technologies
Field of study
  • Technological process management

Publications

Publications (31)
Article
Full-text available
In the modern era, each Internet user leaves enormous amounts of auxiliary digital residuals (footprints) by using a variety of on-line services. All this data is already collected and stored for many years. In recent works, it was demonstrated that it's possible to apply simple machine learning methods to analyze collected digital footprints and t...
Chapter
Cities were built to facilitate trade and protection. A smart city environment can be considered a product of the evolutionary process built around the coevolution of multiple autonomous artificial intelligent agents (AAIAs). The central concept of a genetic algorithm is a population of candidate solutions that evolve to solve optimisation problems...
Experiment Findings
Full-text available
The goal of the controller is to apply the force to the cart keeping two poles balanced as long as possible. At the same time, the cart should stay on track within the defined boundaries. As with single-pole balancing problem discussed before the control strategy can be defined as an avoidance control problem. Which means that the controller must m...
Experiment Findings
Full-text available
The single-pole balancer (a.k.a. inverted pendulum) is an unstable pendulum that has its center of mass above its pivot point. It can be stabilized by applying external forces under control of a specialized system that monitors the angle of the pole and moves the pivot point horizontally back and forth under the center of mass as it starts to fall....
Experiment Findings
Full-text available
The XOR problem solving is a classic computer science experiment in the field of reinforcement learning, which can not be solved without introducing non-linear execution to the solver algorithm. The two inputs to the XOR solver must be combined at some hidden unit, as opposed to only at the output node, because there is no function over a linear c...
Preprint
Full-text available
In this paper, we look at how Artificial Swarm Intelligence can evolve using evolutionary algorithms that try to minimize the sensory surprise of the system. We will show how to apply the free-energy principle, borrowed from statistical physics, to quantitatively describe the optimization method (sensory surprise minimization), which can be used to...
Preprint
Full-text available
In this paper, we look at how Artificial Swarm Intelligence can evolve using evolutionary algorithms that try to minimize the sensory surprise of the system. We will show how to apply the free-energy principle, borrowed from statistical physics, to quantitatively describe the optimization method (sensory surprise minimization), which can be used to...
Preprint
Full-text available
Search for Novelty is an universal method of biological life evolution. We decided to apply it for Autonomous Artificial Intelligent Agents breeding using Neuro-Evolution algorithm to conduct evolutionary process.
Article
Full-text available
This article presents how to build and train Artificial Neural Networks by NEAT algorithm. It will consider weakness of current Gradient Descent based training methods and shows a way to improve it. (Original article was published by author at Medium: https://medium.com/@io42/neuroevolution-evolving-artificial-neural-networks-topology-from-the-scra...
Preprint
Full-text available
The NeuroEvolution (NE) is an artificial evolution of Neural Networks (NN) using genetic algorithms in order to find optimal NN parameters and topology. NeuroEvolution of NN may assume search for optimal weights of connections between NN nodes as well as search for optimal topology of resulting NN. The NEAT algorithm implemented in this work do se...
Presentation
Full-text available
The presentation of research paper: Applying Deep Machine Learning for Psycho-Demographic Profiling of Internet Users using O.C.E.A.N. Model of Personality
Article
Full-text available
In this article we present the results of our research related to the study of correlations between specific visual stimulation and the elicited brain's electro-physiological response collected by EEG sensors from a group of participants. We will look at how the various characteristics of visual stimulation affect the measured electro-physiological...
Patent
Full-text available
Method for data exchange is characterized with preparation of files with information, advertisement or entertainment content suitable for loading to mobile device, with preparation of server of order of file distribution, loading files from server to computer connected to module of wireless data exchange, with placement of computer with module for...

Projects

Projects (7)
Project
The main goal is to evaluate the performance and efficiency of various neuroevolution libraries against well-known standard benchmarks. An additional goal is to define a performance metric allowing to compare different libraries that implement the NEAT algorithm in terms of its evolutionary and computational efficiency.
Project
This project is about writing the book describing popular Neuroevolution-based genetic algorithms with practical examples based on the available Python libraries. This book is intended to serve as a practical guide of how to develop a necessary mindset and skills to apply neuroevolution-based algorithms to solve real-world tasks. The reader will learn the key concepts and methods of neuroevolution by writing code with Python programming language and get hands-on experience with popular Python libraries.
Project
In this work we examine how open-ended evolution strategy can be simulated by Evolved Elastic Artificial Neural Network architecture implemented using ES HyperNEAT algorithm. The open-ended evolution system assumes endless ability to generate novel creative outcomes, which sometimes lead to finding an optimal solutions for a specific practical task. The main inspiration is taken from the nature itself, where billions of years of evolution demonstrated endless creativity in creation of zillions of biological forms suitable for a widest variety of environments, and this evolutionary search process is still on the go. We believe that open-endedness of evolutionary dynamics implemented using evolutionary computational algorithms can lead to major breakthroughs in creation of Artificial General Intelligent systems capable of life-time learning and common-sense reasoning. Through inherent elastic rules and neuro-modulation mechanisms, the neural network modules of the AGI systems will be able to continuously adjust the learned weights of synaptic connections and even modify the network topology in response to changes in the environment or to the internal stimuli caused by novel knowledge assimilation through the self-reflection. In our experiment we will explore the evolution of maze solving agents which coevolve with maze generating agents. Within this process the populations of maze generators will compete against maze solvers, providing new evolutionary opportunities for innovation in both populations. Thus, the algorithm will continually generate novel and increasingly complex artifacts simulating near open-ended evolutionary process. One of the interesting implications of such an experiment is that open-ended coevolution can be applied to any coevolving pairs within the same basic setup. For example, we can imagine that Artificial agent’s body will coevolve with it’s brain which will know how to control the evolved body. By modifying modulating environmental signals it will be possible to evolve variety of physical body configurations optimally suitable for specific environments and with optimal control circuits (low brain) evolved by the same process.