-
Cayley graphs and G-graphs of Gyro-groups
Authors:
Neda Moradi,
Gholam Hossein Fath-Tabara,
Alain Bretto
Abstract:
In this paper the structure of the Cayley graphs and G-graphs of some gyro-groups are studied and some properties of them will be proved. Moreover we review some special gyro-groups including: gyro-commutative gyrogroups, dihedral gyro-groups and dihedralized gyro-groups. Then we try to establish some important properties of their associated G-graphs. Finally we will examine the symmetry of Cayley…
▽ More
In this paper the structure of the Cayley graphs and G-graphs of some gyro-groups are studied and some properties of them will be proved. Moreover we review some special gyro-groups including: gyro-commutative gyrogroups, dihedral gyro-groups and dihedralized gyro-groups. Then we try to establish some important properties of their associated G-graphs. Finally we will examine the symmetry of Cayley graphs and G-graphs of some gyro-groups.
△ Less
Submitted 30 April, 2024;
originally announced April 2024.
-
Feed-Forward Neural Networks as a Mixed-Integer Program
Authors:
Navid Aftabi,
Nima Moradi,
Fatemeh Mahroo
Abstract:
Deep neural networks (DNNs) are widely studied in various applications. A DNN consists of layers of neurons that compute affine combinations, apply nonlinear operations, and produce corresponding activations. The rectified linear unit (ReLU) is a typical nonlinear operator, outputting the max of its input and zero. In scenarios like max pooling, where multiple input values are involved, a fixed-pa…
▽ More
Deep neural networks (DNNs) are widely studied in various applications. A DNN consists of layers of neurons that compute affine combinations, apply nonlinear operations, and produce corresponding activations. The rectified linear unit (ReLU) is a typical nonlinear operator, outputting the max of its input and zero. In scenarios like max pooling, where multiple input values are involved, a fixed-parameter DNN can be modeled as a mixed-integer program (MIP). This formulation, with continuous variables representing unit outputs and binary variables for ReLU activation, finds applications across diverse domains. This study explores the formulation of trained ReLU neurons as MIP and applies MIP models for training neural networks (NNs). Specifically, it investigates interactions between MIP techniques and various NN architectures, including binary DNNs (employing step activation functions) and binarized DNNs (with weights and activations limited to $-1,0,+1$). The research focuses on training and evaluating proposed approaches through experiments on handwritten digit classification models. The comparative study assesses the performance of trained ReLU NNs, shedding light on the effectiveness of MIP formulations in enhancing training processes for NNs.
△ Less
Submitted 8 February, 2024;
originally announced February 2024.
-
Quantum Decoherence Timescales for Ionic Superposition States in Ion Channels
Authors:
V. Salari,
N. Moradi,
F. Fazileh,
F. Shahbazi
Abstract:
There are many controversial and challenging discussions about quantum effects in microscopic structures in neurons of the human brain. The challenge is mainly because of quick decoherence of quantum states due to hot, wet and noisy environment of the brain which forbids long life coherence for brain processing. Despite these critical discussions, there are only a few number of published papers ab…
▽ More
There are many controversial and challenging discussions about quantum effects in microscopic structures in neurons of the human brain. The challenge is mainly because of quick decoherence of quantum states due to hot, wet and noisy environment of the brain which forbids long life coherence for brain processing. Despite these critical discussions, there are only a few number of published papers about numerical aspects of decoherence in neurons. Perhaps the most important issue is offered by Max Tegmark who has calculated decoherence times for the systems of "ions" and "microtubules" in neurons of the brain. In fact, Tegmark did not consider ion channels which are responsible for ions displacement through the membrane and are the building blocks of electrical membrane signals in the nervous system. Here, we would like to re-investigate decoherence times for ionic superposition states by using the data obtained via molecular dynamics simulations. Our main approach is according to what Tegmark has used before. In fact, Tegmark didn't consider the ion channel structure and his estimates are only simple approximations. In this paper, we focus on the small nano-scale part of KcsA ion channels which is called "selectivity filter" and has a key role in the operation of an ion channel. Our results for superposition states of potassium ions indicate that decoherence times are in the order of picoseconds which are 10-100 million times bigger than the order calculated by Tegmark. This decoherence time is still not enough for cognitive processing in the brain, however it can be adequate for quantum states of cooled ions in the filter to leave their quantum traces on the filter and action potentials.
△ Less
Submitted 23 October, 2014;
originally announced October 2014.
-
Selectivity filter gate versus voltage-sensitive gate: A study of quantum probabilities in the Hodgkin-Huxley equation
Authors:
Narges Moradi,
Felix Scholkmann,
Vahid Salari
Abstract:
The Hodgkin-Huxley (HH) model is a powerful model to explain different aspects of spike generation in excitable cells. However, the HH model was proposed in 1952 when the real structure of the ion channel was unknown. It is now common knowledge that in many ion-channel proteins the flow of ions through the pore is governed by a gate, comprising a so-called selectivity filter inside the ion channel…
▽ More
The Hodgkin-Huxley (HH) model is a powerful model to explain different aspects of spike generation in excitable cells. However, the HH model was proposed in 1952 when the real structure of the ion channel was unknown. It is now common knowledge that in many ion-channel proteins the flow of ions through the pore is governed by a gate, comprising a so-called selectivity filter inside the ion channel, which can be controlled by electrical interactions. The selectivity filter is believed to be responsible for the selection and fast conduction of particular ions across the membrane of an excitable cell. Other (generally larger) parts of the molecule such as the pore-domain gate control the access of ions to the channel protein. In fact, two types of gates are considered here for ion channels: the external gate, which is the voltage sensitive gate, and the internal gate which is the selectivity filter gate (SFG). Some quantum effects are to expected in the SFG due to its small dimensions, which may play an important role in the operation of an ion channel. Here, we examine parameters in a generalized model of HH to see whether any parameter affects the spike generation. Our results indicate that the previously suggested semi-quantum-classical equation proposed by Bernroider and Summhammer (BS) agrees strongly with the HH equation under different conditions and may even provide a better explanation in some cases. We conclude that the BS model can refine the classical HH model substantially.
△ Less
Submitted 5 October, 2014;
originally announced October 2014.
-
Accelerated kinetic over-relaxation for dynamic energy minimization in lattice many-body problems
Authors:
Nasrollah Moradi,
Andreas Greiner,
Simone Melchionna,
Francesco Rao,
Sauro Succi
Abstract:
A kinetic over-relaxation minimizer, based on a lattice version of the Boltzmann equation is presented. The method is validated against standard Metropolis Monte Carlo, and proves very effective in attaining (global) minima of classical pair potentials, involving solid body rotations. Linear scaling of the computational time to minimum with the system size is demonstrated.
A kinetic over-relaxation minimizer, based on a lattice version of the Boltzmann equation is presented. The method is validated against standard Metropolis Monte Carlo, and proves very effective in attaining (global) minima of classical pair potentials, involving solid body rotations. Linear scaling of the computational time to minimum with the system size is demonstrated.
△ Less
Submitted 10 March, 2014;
originally announced March 2014.
-
Dynamics of cylindrical droplets on flat substrate: Lattice Boltzmann modeling versus simple analytic models
Authors:
Nasrollah Moradi,
Fathollah Varnik,
Ingo Steinbach
Abstract:
The steady state motion of cylindrical droplets under the action of external body force is investigated both theoretically and via lattice Boltzmann simulation. As long as the shape-invariance of droplet is maintained, the droplet's center-of-mass velocity linearly scales with both the force density and the square of droplet radius. However, a non-linear behavior appears as the droplet deformation…
▽ More
The steady state motion of cylindrical droplets under the action of external body force is investigated both theoretically and via lattice Boltzmann simulation. As long as the shape-invariance of droplet is maintained, the droplet's center-of-mass velocity linearly scales with both the force density and the square of droplet radius. However, a non-linear behavior appears as the droplet deformation becomes significant. This deformation is associated with the drop elongation occurring at sufficiently high external forcing. Yet, independent of either the force density or the droplet size, the center-of-mass velocity is found to be linear in terms of the inverse of dynamic viscosity. In addition, it is shown that the energy is mainly dissipated in a region near the substrate particularly close to the three phase contact line. The total viscous dissipation is found to be proportional to both the square of force density and the inverse of dynamic viscosity. Moreover, the dependence of the center-of-mass velocity on the equilibrium contact angle is investigated. A simple analytic model is provided reproducing the observed behavior.
△ Less
Submitted 10 February, 2011;
originally announced February 2011.
-
Roughness gradient induced spontaneous motion of droplets on hydrophobic surfaces: A lattice Boltzmann study
Authors:
Nasrollah Moradi,
fathollah Varnik,
Ingo Steinbach
Abstract:
The effect of a step wise change in the pillar density on the dynamics of droplets is investigated via three-dimensional lattice Boltzmann simulations. For the same pillar density gradient but different pillar arrangements, both motion over the gradient zone as well as complete arrest are observed. In the moving case, the droplet velocity scales approximately linearly with the texture gradient.…
▽ More
The effect of a step wise change in the pillar density on the dynamics of droplets is investigated via three-dimensional lattice Boltzmann simulations. For the same pillar density gradient but different pillar arrangements, both motion over the gradient zone as well as complete arrest are observed. In the moving case, the droplet velocity scales approximately linearly with the texture gradient. A simple model is provided reproducing the observed linear behavior. The model also predicts a linear dependence of droplet velocity on surface tension. This prediction is clearly confirmed via our computer simulations for a wide range of surface tensions.
△ Less
Submitted 16 October, 2009;
originally announced October 2009.
