Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

Facta universitatis - series Electronics and Energetics

Inductive learning of quantum behaviors

Martin Lukáč, Marek A. Perkowski

Published Jan 1, 2007• 10 citations
Publisher

Abstract

In this paper studied are new concepts of robotic behaviors - deterministic and quantum probabilistic. In contrast to classical circuits, the quantum circuit can realize both of these behaviors. When applied to a robot, a quantum circuit controller realizes what we call quantum robot behaviors. We use automated methods to synthesize quantum behaviors (circuits) from the examples (examples are cares of the quantum truth table). The don't knows (minterms not given as examples) are then converted not only to deterministic cares as in the classical learning, but also to output values generated with various probabilities. The Occam Razor principle, fundamental to inductive learning, is satisfied in this approach by seeking circuits of reduced complexity. This is illustrated by the synthesis of single output quantum circuits, as we extended the logic synthesis approach to Inductive Machine Learning for the case of learning quantum circuits from behavioral examples.

DOI
10.2298/fuee0703561l

Related papers

Evolutionary approach to quantum symbolic logic synthesis

Martin Lukac, Marek A. Perkowski

In this paper we present an evolutionary approach to the quantum symbolic logic synthesis. We use a genetic algorithm to synthesize quantum circuits from examples, allowing to synthesize functions that are both completely and incompletely specified. The symbolic synthesis is implemented in the GA so as to verify our approach. The Occam Razor principle, fundamental to inductive learning as well as to logic synthesis, is satisfied in this approach by seeking circuits of reduced complexity. The GA is tested on a set of benchmark functions representing single output quantum circuits as well as multiple entangled-qubit state generators.

IEEE Congress on Evolutionary Computation (CEC)•Jun 1, 2008•11 citations

Quantum Motions and Emotions for a Humanoid Robot Actor

Richard Deng, Yuchen Huang, Marek A. Perkowski

No quantum robot exists in the world today, meaning a real physical robot, controlled by a “quantum brain”. In this paper a humanoid robot was programmed based on quantum circuits, in contrast to classical animation methods or classical logic circuit robot controllers. The robot actor executes deterministic, probabilistic and quantum probabilistic behaviors which makes him more entertaining to the robot theatre audience.

IEEE International Symposium on Multiple-Valued Logic (ISMVL)•May 1, 2021•10 citations

Quantum Computing for Designing Behavioural Model and Quantum Machine Learning on a Humanoid Robot

Jacob Viertel, Mohamed Aburaia

Today's robots are controlled with classical computers that make decisions based on algorithms designed for a classical computing architecture.However, the amount of data increases continuously -a fact which pushes classical computers that control these robots to their limits.In comparison to classical computing, fundamentally different concepts of physics are the basis of quantum computing and thereby quantum computing has the potential to solve specific computational problems more efficiently.In addition to the aforementioned complications, the subgroup of robots that are subject to the research of this paper, namely humanoid robots, not only need more computational capacity but also require an algorithmic design based on a behavioral model which avoids causing antipathy when interacting with human beings.Using a real quantum computer with quantum phenomena such as superposition, entanglement, and quantum parallelism, new behavioral models and algorithms can be designed to improve human robot interaction and speed up machine learning processes.This paper promotes a hybrid-controlled robot using the combination of a classical computer-and quantum computer.With four quantum circuits to define a behavioral model and three different quantum machine learning algorithms we show the advantage of the hybrid-controlled robot.

Annals of DAAAM for ... & proceedings of the ... International DAAAM Symposium•Jan 1, 2021•0 citations

Quantum-Inspired Probabilistic Circuit Synthesis

Jincheng Zhang

This paper explores a novel approach to neural circuit synthesis, drawing inspiration from the principles of quantum mechanics. Traditional neural network optimization often relies on deterministic algorithms, which can be susceptible to local optima and limited exploration of the solution space. This work proposes a methodology termed "Quantum-Inspired Probabilistic Circuit Synthesis" that leverages concepts like superposition and interference to introduce probabilistic pathways within a neural network architecture. The core idea is to generate a network where connections between nodes are not fixed but governed by probabilities, allowing for a more diverse and robust exploration of the circuit's behavior. We formulate a synthesis process based on probabilistic gate selection and connection weighting, mimicking quantum interference to constructively reinforce promising pathways and destructively cancel unfavorable ones. This approach aims to significantly improve the efficiency and robustness of neural circuit design, offering a potentially superior alternative to conventional deterministic methods. The theoretical framework is presented, along with a description of the core algorithmic components, and a discussion of the potential benefits and challenges associated with this new paradigm.

•Sep 3, 2026•0 citations

AutoQC: Automated Synthesis of Quantum Circuits Using Neural Network

Murakami Kentaro, Jianjun Zhao

While the ability to build quantum computers is improving dramatically, developing quantum algorithms is limited and relies on human insight and ingenuity. Although a number of quantum programming languages have been developed, it is challenging for software developers who are not familiar with quantum computing to learn and use these languages. It is, therefore, necessary to develop tools to support developing new quantum algorithms and programs automatically. This paper proposes AutoQC, an approach to automatically synthesizing quantum circuits using the neural network from input and output pairs. We consider a quantum circuit a sequence of quantum gates and synthesize a quantum circuit probabilistically by prioritizing with a neural network at each step. The experimental results highlight the ability of AutoQC to synthesize some essential quantum circuits at a lower cost.

arXiv:2210.02766•Oct 6, 2022•1 citation

Realization of Deterministic Quantum Circuits for Non-Deterministic or Incompletely Specified Quantum State Machines

Manjith Kumar, Marek A. Perkowski

Abstract In classical logic design, there are machine learning methods based on converting a set of input-output traces to non-deterministic automata that are then converted to deterministic automata and synthesized using logic gates. This approach has not yet been extended to quantum automata. In this paper, we present a method to convert a set of input-output traces to a non-deterministic automaton, which is then converted to an incompletely specified multi-output Boolean function. The existing logic synthesis approaches for designing quantum circuits are insufficient to handle incompletely specified functions. So, we present a novel algorithm to synthesize logic functions with don’t cares using permutative quantum gates. The original MMD (D.M.Miller, D. Maslov, and G.W.Dueck) algorithm minimizes only completely specified reversible functions using cascades of reversible gates. In this paper, this algorithm is modified to allow for the inclusion of don’t cares within the given function’s truth table (reversible or irreversible). The distinguishing property of our presented algorithm QAS, is that it does not add any ancilla qubits if it is not necessary. This algorithm solves the problem of synthesizing deterministic and non-deterministic quantum state machines, both completely specified and incompletely specified.

Quantum Information and Computation•Dec 1, 2025•0 citations
In this paper studied are new concepts of robotic behaviors - deterministic and quantum probabilistic. In contrast to classical circuits, the quantum circuit can realize both of these behaviors. When applied to a robot, a quantum circuit controller realizes what we call quantum robot behaviors. We use automated methods to synthesize quantum behaviors (circuits) from the examples (examples are cares of the quantum truth table). The don't knows (minterms not given as examples) are then converted not only to deterministic cares as in the classical learning, but also to output values generated with various probabilities. The Occam Razor principle, fundamental to inductive learning, is satisfied in this approach by seeking circuits of reduced complexity. This is illustrated by the synthesis of single output quantum circuits, as we extended the logic synthesis approach to Inductive Machine Learning for the case of learning quantum circuits from behavioral examples.
In this paper we present an evolutionary approach to the quantum symbolic logic synthesis. We use a genetic algorithm to synthesize quantum circuits from examples, allowing to synthesize functions that are both completely and incompletely specified. The symbolic synthesis is implemented in the GA so as to verify our approach. The Occam Razor principle, fundamental to inductive learning as well as to logic synthesis, is satisfied in this approach by seeking circuits of reduced complexity. The GA is tested on a set of benchmark functions representing single output quantum circuits as well as multiple entangled-qubit state generators.
Today's robots are controlled with classical computers that make decisions based on algorithms designed for a classical computing architecture.However, the amount of data increases continuously -a fact which pushes classical computers that control these robots to their limits.In comparison to classical computing, fundamentally different concepts of physics are the basis of quantum computing and thereby quantum computing has the potential to solve specific computational problems more efficiently.In addition to the aforementioned complications, the subgroup of robots that are subject to the research of this paper, namely humanoid robots, not only need more computational capacity but also require an algorithmic design based on a behavioral model which avoids causing antipathy when interacting with human beings.Using a real quantum computer with quantum phenomena such as superposition, entanglement, and quantum parallelism, new behavioral models and algorithms can be designed to improve human robot interaction and speed up machine learning processes.This paper promotes a hybrid-controlled robot using the combination of a classical computer-and quantum computer.With four quantum circuits to define a behavioral model and three different quantum machine learning algorithms we show the advantage of the hybrid-controlled robot.
This paper explores a novel approach to neural circuit synthesis, drawing inspiration from the principles of quantum mechanics. Traditional neural network optimization often relies on deterministic algorithms, which can be susceptible to local optima and limited exploration of the solution space. This work proposes a methodology termed "Quantum-Inspired Probabilistic Circuit Synthesis" that leverages concepts like superposition and interference to introduce probabilistic pathways within a neural network architecture. The core idea is to generate a network where connections between nodes are not fixed but governed by probabilities, allowing for a more diverse and robust exploration of the circuit's behavior. We formulate a synthesis process based on probabilistic gate selection and connection weighting, mimicking quantum interference to constructively reinforce promising pathways and destructively cancel unfavorable ones. This approach aims to significantly improve the efficiency and robustness of neural circuit design, offering a potentially superior alternative to conventional deterministic methods. The theoretical framework is presented, along with a description of the core algorithmic components, and a discussion of the potential benefits and challenges associated with this new paradigm.
While the ability to build quantum computers is improving dramatically, developing quantum algorithms is limited and relies on human insight and ingenuity. Although a number of quantum programming languages have been developed, it is challenging for software developers who are not familiar with quantum computing to learn and use these languages. It is, therefore, necessary to develop tools to support developing new quantum algorithms and programs automatically. This paper proposes AutoQC, an approach to automatically synthesizing quantum circuits using the neural network from input and output pairs. We consider a quantum circuit a sequence of quantum gates and synthesize a quantum circuit probabilistically by prioritizing with a neural network at each step. The experimental results highlight the ability of AutoQC to synthesize some essential quantum circuits at a lower cost.
Abstract In classical logic design, there are machine learning methods based on converting a set of input-output traces to non-deterministic automata that are then converted to deterministic automata and synthesized using logic gates. This approach has not yet been extended to quantum automata. In this paper, we present a method to convert a set of input-output traces to a non-deterministic automaton, which is then converted to an incompletely specified multi-output Boolean function. The existing logic synthesis approaches for designing quantum circuits are insufficient to handle incompletely specified functions. So, we present a novel algorithm to synthesize logic functions with don’t cares using permutative quantum gates. The original MMD (D.M.Miller, D. Maslov, and G.W.Dueck) algorithm minimizes only completely specified reversible functions using cascades of reversible gates. In this paper, this algorithm is modified to allow for the inclusion of don’t cares within the given function’s truth table (reversible or irreversible). The distinguishing property of our presented algorithm QAS, is that it does not add any ancilla qubits if it is not necessary. This algorithm solves the problem of synthesizing deterministic and non-deterministic quantum state machines, both completely specified and incompletely specified.