Showing 1–15 of 15 results for author: Gaur, D

Reflection and Transmission of Airy Pulse from Controllable Periodic Temporal Boundary

Authors: Deependra Singh Gaur, Akhilesh Kumar Mishra

Abstract: We numerically investigate the interaction between two Airy pulses propagating at different wavelengths. The periodically varying peak intensity of the soliton that emerges from stronger Airy pulse (pump pulse) leads to the formation of periodic temporal boundary. The relatively weaker Airy pulse (probe pulse) on interaction with this boundary gets partially reflected as well as transmitted. As a… ▽ More We numerically investigate the interaction between two Airy pulses propagating at different wavelengths. The periodically varying peak intensity of the soliton that emerges from stronger Airy pulse (pump pulse) leads to the formation of periodic temporal boundary. The relatively weaker Airy pulse (probe pulse) on interaction with this boundary gets partially reflected as well as transmitted. As a result, the probe pulse spectrum splits into two parts -- the reflected pulse spectrum undergoes red shift while transmitted pulse exhibits blue shift. The probe pulse witneses maximum reflection when point of interaction lies on the intensity maxima of the emergent soliton from pump Airy pulse. On the other hand, maximum transmission occurs when probe Airy pulse interacts at the intensity minima of the soliton. The reflection and transmission processes can be manipulated by tuning the time delay between pump and probe Airy pulses. In case of sufficiently intense pump pulse, the temporal boundary mimics the artificial optical event horizon, and the weak probe Airy pulse is completely reflected. This phenomenon is equivalent to the temporal version of total internal reflection. The time delay and truncation parameters play a vital role in the interaction. The results of the study hold potential applications in optical manipulation and temporal waveguiding. △ Less

Submitted 1 January, 2024; originally announced January 2024.

A Hybrid Quantum Algorithm for Load Flow

Authors: David Neufeld, Sajad Fathi Hafshejani, Daya Gaur, Robert Benkoczi

Abstract: The goal of the load flow study is to ensure that electrical power is delivered efficiently and reliably to end-users while maintaining the stability and security of the power system. Newton-Raphson is a numerical method used widely for load flow analysis. One of the most computationally expensive steps in this method is an equation-solving step. We propose to replace this step with HHL, a quantum… ▽ More The goal of the load flow study is to ensure that electrical power is delivered efficiently and reliably to end-users while maintaining the stability and security of the power system. Newton-Raphson is a numerical method used widely for load flow analysis. One of the most computationally expensive steps in this method is an equation-solving step. We propose to replace this step with HHL, a quantum algorithm for solving linear systems of equations. HHL is exponentially faster, but with caveats. In this study, a hybrid quantum algorithm is proposed for solving load flow. The Newton-Raphson method is used as a benchmark to compare the performance of the hybrid quantum algorithm. Although the simulation of the hybrid quantum algorithm takes much time, these preliminary results are encouraging and point to the potential for the use of quantum algorithms to develop hybrid quantum algorithms for load flow analysis and related problems. △ Less

Submitted 27 June, 2024; v1 submitted 30 October, 2023; originally announced October 2023.

Journal ref: Power, Energy and Electrical Engineering, Volume 54, 2024, Pages 589-600

Quench dynamics across the MI-SF quantum phase transition with cluster mean field theory

Authors: Deepak Gaur, Hrushikesh Sable, D. Angom

Abstract: In this work, we study the quench dynamics of quantum phases of ultracold neutral bosons trapped in optical lattices. We investigate the validity of the Kibble-Zurek (KZ) scaling laws with the single-site Gutzwiller mean-field (SGMF) and cluster Gutzwiller mean-field (CGMF) theory. With CGMF, we note the evolution of the dynamical wavefunction in the ``impulse" regime of the Kibble-Zurek mechanism… ▽ More In this work, we study the quench dynamics of quantum phases of ultracold neutral bosons trapped in optical lattices. We investigate the validity of the Kibble-Zurek (KZ) scaling laws with the single-site Gutzwiller mean-field (SGMF) and cluster Gutzwiller mean-field (CGMF) theory. With CGMF, we note the evolution of the dynamical wavefunction in the ``impulse" regime of the Kibble-Zurek mechanism. We obtain the power law scalings for the crossover time and defect density with the quench rate predicted by KZ scaling laws. The critical exponents obtained from dynamics are close to their equilibrium values. Furthermore, it is observed that the obtained dynamical critical exponent $z$ improves towards the equilibrium value with increasing cluster sizes in CGMF. △ Less

Submitted 12 September, 2023; originally announced September 2023.

Comments: 8 pages, 13 figures, Comments are most welcome

Exact-diagonalization method for soft-core bosons in optical lattices using hierarchical wavefunctions

Authors: Deepak Gaur, Hrushikesh Sable, D. Angom

Abstract: In this work, we describe a new technique for numerical exact-diagonalization. The method is particularly suitable for cold bosonic atoms in optical lattices, in which multiple atoms can occupy a lattice site. We describe the use of the method for Bose-Hubbard model as an example, however, the method is general and can be applied to other lattice models. The proposed numerical technique focuses in… ▽ More In this work, we describe a new technique for numerical exact-diagonalization. The method is particularly suitable for cold bosonic atoms in optical lattices, in which multiple atoms can occupy a lattice site. We describe the use of the method for Bose-Hubbard model as an example, however, the method is general and can be applied to other lattice models. The proposed numerical technique focuses in detail on how to construct the basis states as a hierarchy of wavefunctions. Starting from single-site Fock states we construct the basis set in terms of row-states and cluster-states. This simplifies the application of constraints and calculation of the Hamiltonian matrix. Each step of the method can be parallelized to accelerate the computation. In addition, we have illustrated the computation of the spatial bipartite entanglement entropy in the correlated $ν=1/2$ fractional quantum Hall state. △ Less

Submitted 14 August, 2023; originally announced August 2023.

Comments: 11 pages, 6 figures, Comments are most welcome

Binary Orthogonal Non-negative Matrix Factorization

Authors: S. Fathi Hafshejani, D. Gaur, S. Hossain, R. Benkoczi

Abstract: We propose a method for computing binary orthogonal non-negative matrix factorization (BONMF) for clustering and classification. The method is tested on several representative real-world data sets. The numerical results confirm that the method has improved accuracy compared to the related techniques. The proposed method is fast for training and classification and space efficient. We propose a method for computing binary orthogonal non-negative matrix factorization (BONMF) for clustering and classification. The method is tested on several representative real-world data sets. The numerical results confirm that the method has improved accuracy compared to the related techniques. The proposed method is fast for training and classification and space efficient. △ Less

Submitted 19 October, 2022; originally announced October 2022.

Fine-grained domain counting and percolation analysis in 2D lattice systems with linked-lists

Authors: Hrushikesh Sable, Deepak Gaur, D. Angom

Abstract: We present a fine-grained approach to identify clusters and perform percolation analysis in a 2D lattice system. In our approach, we develop an algorithm based on the linked-list data structure whereby the members of a cluster are nodes of a path. This path is mapped to a linked-list. This approach facilitates unique cluster labeling in a lattice with a single scan. We use the algorithm to determi… ▽ More We present a fine-grained approach to identify clusters and perform percolation analysis in a 2D lattice system. In our approach, we develop an algorithm based on the linked-list data structure whereby the members of a cluster are nodes of a path. This path is mapped to a linked-list. This approach facilitates unique cluster labeling in a lattice with a single scan. We use the algorithm to determine the critical exponent in the quench dynamics from the Mott insulator to the superfluid phase of bosons in 2D square optical lattices. The results obtained are consistent with the Kibble-Zurek mechanism. We also employ the algorithm to compute the correlation length using definitions based on percolation theory and use it to identify the quantum critical point of the Bose Glass to superfluid transition in the disordered 2D square optical lattices. In addition, we compute the critical exponent $ν$ which quantify the divergence of the correlation length $ξ$ across the phase transition and the fractal dimension of the hulls of the superfluid clusters. △ Less

Submitted 26 October, 2023; v1 submitted 31 December, 2021; originally announced December 2021.

Comments: 15 pages, 15 figures, Published in the October issue of Physical Review E (Vol. 108, No. 4)

Journal ref: Phys. Rev. E 108, 045307 (2023)

Conditions for Advantageous Quantum Bitcoin Mining

Authors: Robert R. Nerem, Daya R. Gaur

Abstract: Our aim is to determine conditions for quantum computing technology to give rise to security risks associated with quantum Bitcoin mining. Specifically, we determine the speed and energy efficiency a quantum computer needs to offer an advantage over classical mining. We analyze the setting in which the Bitcoin network is entirely classical except for a single quantum miner who has small hash rate… ▽ More Our aim is to determine conditions for quantum computing technology to give rise to security risks associated with quantum Bitcoin mining. Specifically, we determine the speed and energy efficiency a quantum computer needs to offer an advantage over classical mining. We analyze the setting in which the Bitcoin network is entirely classical except for a single quantum miner who has small hash rate compared to that of the network. We develop a closed-form approximation for the probability that the quantum miner successfully mines a block, with this probability dependent on the number of Grover iterations the quantum miner applies before making a measurement. Next, we show that, for a quantum miner that is "peaceful", this success probability is maximized if the quantum miner applies Grover iterations for 16 minutes before measuring, which is surprising as the network mines blocks every 10 minutes on average. Using this optimal mining procedure, we show that the quantum miner outperforms a classical computer in efficiency (cost per block) if the condition $Q < Crb$ is satisfied, where $Q$ is the cost of a Grover iteration, $C$ is the cost of a classical hash, $r$ is the quantum miner's speed in Grover iterations per second, and $b$ is a factor that attains its maximum if the quantum miner uses our optimal mining procedure. This condition lays the foundation for determining when quantum mining, and the known security risks associated with it, will arise. △ Less

Submitted 2 October, 2021; originally announced October 2021.

Comments: 16 pages, 2 figures

Barzilai and Borwein conjugate gradient method equipped with a non-monotone line search technique and its application on non-negative matrix factorization

Authors: Sajad Fathi Hafshejani, Daya Gaur, Shahadat Hossain, Robert Benkoczi

Abstract: In this paper, we propose a new non-monotone conjugate gradient method for solving unconstrained nonlinear optimization problems. We first modify the non-monotone line search method by introducing a new trigonometric function to calculate the non-monotone parameter, which plays an essential role in the algorithm's efficiency. Then, we apply a convex combination of the Barzilai-Borwein method for c… ▽ More In this paper, we propose a new non-monotone conjugate gradient method for solving unconstrained nonlinear optimization problems. We first modify the non-monotone line search method by introducing a new trigonometric function to calculate the non-monotone parameter, which plays an essential role in the algorithm's efficiency. Then, we apply a convex combination of the Barzilai-Borwein method for calculating the value of step size in each iteration. Under some suitable assumptions, we prove that the new algorithm has the global convergence property. The efficiency and effectiveness of the proposed method are determined in practice by applying the algorithm to some standard test problems and non-negative matrix factorization problems. △ Less

Submitted 12 September, 2021; originally announced September 2021.

Journal ref: Optimization for Machine Learning Workshop @ NeurIPS 2021

Soliton Shedding from Airy Pulses in a Highly Dispersive and Nonlinear Medium

Authors: Deependra Singh Gaur, Ankit Purohit, Akhilesh Kumar Mishra

Abstract: We present a numerical investigation of the propagation dynamics of a truncated Airy pulse in a highly dispersive and nonlinear medium by employing the split-step Fourier transform method and look, in particular, into the effects of fourth order dispersion (FOD) and cubic-quintic-septic nonlinearity on pulse evolution. Presence of FOD cancels the Airy pulse self-acceleration along with eclipsing t… ▽ More We present a numerical investigation of the propagation dynamics of a truncated Airy pulse in a highly dispersive and nonlinear medium by employing the split-step Fourier transform method and look, in particular, into the effects of fourth order dispersion (FOD) and cubic-quintic-septic nonlinearity on pulse evolution. Presence of FOD cancels the Airy pulse self-acceleration along with eclipsing the oscillatory tail during propagation in the linear regime. Further, we observe soliton shedding at low input pulse power in the presence of cubic and quintic nonlinearity and negative FOD. The emergent soliton exhibits temporal shift and the direction and the extent of the shift depend upon the strengths of cubic and quintic nonlinearities. In the presence of anomalous group velocity dispersion (GVD) with negative FOD, soliton shedding is observed at relatively high input pulse power. The strengths of GVD and nonlinearity play a vital role in the temporal shifting of emergent soliton. Furthermore, we have explored the effects of septic nonlinearity on soliton shedding in different scenarios of nonlinearity and dispersion. △ Less

Submitted 27 October, 2021; v1 submitted 28 July, 2021; originally announced July 2021.

Comments: 11 pages, 12 figures

Journal ref: Journal of the Optical Society of America B (2021)

Dynamics of chirped Airy pulse in a dispersive medium with higher-order nonlinearity

Authors: Ankit Purohit, Deependra Singh Gaur, Akhilesh Kumar Mishra

Abstract: Chirp can control the dynamics of the Airy pulse, making it an essential factor in pulse manipulation. Finite energy chirped Airy pulse (FECAP) has potential applications in underwater optical communication and imaging. Hence, it's critical to study the propagation of FECAP. We present a numerical investigation of the propagation dynamics of a FECAP in a dispersive, and highly nonlinear medium. Th… ▽ More Chirp can control the dynamics of the Airy pulse, making it an essential factor in pulse manipulation. Finite energy chirped Airy pulse (FECAP) has potential applications in underwater optical communication and imaging. Hence, it's critical to study the propagation of FECAP. We present a numerical investigation of the propagation dynamics of a FECAP in a dispersive, and highly nonlinear medium. The nonlinearity under study includes self-phase modulation (SPM), self-steepening (SS), as well as intra-pulse Raman scattering (IRS) terms. We have observed soliton shedding and the chirp parameter is demonstrated to have a considerable impact on the pulse dynamics. In particular, the emergent soliton does not propagate in a straight path instead, depending on the sign of the chirp parameter, it delays or advances in the time. Furthermore, it has been established that the chirp can be employed as an alternate control parameter for the spectral manipulation. The results of our study may have implications in supercontinuum generation and for producing tunable sources. △ Less

Submitted 24 October, 2021; v1 submitted 15 July, 2021; originally announced July 2021.

Comments: 12 pages, 8 figures

Journal ref: Journal of the Optical Society of America B (2021)

Quantum phases of dipolar bosons in multilayer optical lattice

Authors: Soumik Bandyopadhyay, Hrushikesh Sable, Deepak Gaur, Rukmani Bai, Subroto Mukerjee, D. Angom

Abstract: We consider a minimal model to investigate the quantum phases of hardcore, polarized dipolar atoms confined in multilayer optical lattices. The model is a variant of the extended Bose-Hubbard model, which incorporates intralayer repulsion and interlayer attraction between the atoms in nearest-neighbour sites. We study the phases of this model emerging from the competition between the attractive in… ▽ More We consider a minimal model to investigate the quantum phases of hardcore, polarized dipolar atoms confined in multilayer optical lattices. The model is a variant of the extended Bose-Hubbard model, which incorporates intralayer repulsion and interlayer attraction between the atoms in nearest-neighbour sites. We study the phases of this model emerging from the competition between the attractive interlayer interaction and the interlayer hop**. Our results from the analytical and cluster-Gutzwiller mean-field theories reveal that multimer formation occurs in the regime of weak intra and interlayer hop** due to the attractive interaction. In addition, intralayer isotropic repulsive interaction results in the checkerboard ordering of the multimers. This leads to an incompressible checkerboard multimer phase at half-filling. At higher interlayer hop**, the multimers are destabilized to form resonating valence-bond like states. Furthermore, we discuss the effects of thermal fluctuations on the quantum phases of the system. △ Less

Submitted 28 June, 2021; originally announced June 2021.

Comments: 16 pages, 10 figures, comments are welcome

Journal ref: Phys. Rev. A 106, 043301 (2022)

Quantum quench dynamics of tilted dipolar bosons in 2D optical lattices

Authors: Hrushikesh Sable, Deepak Gaur, Soumik Bandyopadhyay, Rejish Nath, D. Angom

Abstract: We investigate the quench dynamics of the dipolar bosons in two dimensional optical lattice of square geometry using the time dependent Gutzwiller method. The system exhibits different density orders like the checkerboard and the striped pattern, depending upon the polarization angle of the dipoles. We quench the hop** parameter across the striped density wave (SDW) to striped supersolid (SSS) p… ▽ More We investigate the quench dynamics of the dipolar bosons in two dimensional optical lattice of square geometry using the time dependent Gutzwiller method. The system exhibits different density orders like the checkerboard and the striped pattern, depending upon the polarization angle of the dipoles. We quench the hop** parameter across the striped density wave (SDW) to striped supersolid (SSS) phase transition, and obtain the scaling laws for the correlation length and topological vortex density, as function of the quench rate. The results are reminiscent of the Kibble-Zurek mechanism (KZM). We also investigate the dynamics from the striped supersolid phase to the checkerboard supersolid phase, obtained by quenching the dipole tilt angle $θ$. This is a first order structural quantum phase transition, and we study the non-equilibrium dynamics from the perspective of the KZM. In particular, we find the number of the domains with checkerboard order follows a power law scaling with the quench rate. This indicates the applicability of the KZM to this first order quantum phase transition. △ Less

Submitted 21 June, 2021; v1 submitted 3 June, 2021; originally announced June 2021.

Comments: 13 pages, 14 figures, revised version, comments are welcome

Training Deep Neural Networks Without Batch Normalization

Authors: Divya Gaur, Joachim Folz, Andreas Dengel

Abstract: Training neural networks is an optimization problem, and finding a decent set of parameters through gradient descent can be a difficult task. A host of techniques has been developed to aid this process before and during the training phase. One of the most important and widely used class of method is normalization. It is generally favorable for neurons to receive inputs that are distributed with ze… ▽ More Training neural networks is an optimization problem, and finding a decent set of parameters through gradient descent can be a difficult task. A host of techniques has been developed to aid this process before and during the training phase. One of the most important and widely used class of method is normalization. It is generally favorable for neurons to receive inputs that are distributed with zero mean and unit variance, so we use statistics about dataset to normalize them before the first layer. However, this property cannot be guaranteed for the intermediate activations inside the network. A widely used method to enforce this property inside the network is batch normalization. It was developed to combat covariate shift inside networks. Empirically it is known to work, but there is a lack of theoretical understanding about its effectiveness and potential drawbacks it might have when used in practice. This work studies batch normalization in detail, while comparing it with other methods such as weight normalization, gradient clip** and dropout. The main purpose of this work is to determine if it is possible to train networks effectively when batch normalization is removed through adaption of the training process. △ Less

Submitted 18 August, 2020; originally announced August 2020.

Segregated quantum phases of dipolar bosonic mixtures in two-dimensional optical lattices

Authors: Rukmani Bai, Deepak Gaur, Hrushikesh Sable, Soumik Bandyopadhyay, K. Suthar, D. Angom

Abstract: We identify the quantum phases in a binary mixture of dipolar bosons in two-dimensional optical lattices. Our study is motivated by the recent experimental realization of binary dipolar condensate mixtures of Er-Dy [Phys. Rev. Lett. 121, 213601 (2018)]. We model the system by using the extended two-species Bose-Hubbard model and calculate the ground-state phase diagrams by using mean-field theory.… ▽ More We identify the quantum phases in a binary mixture of dipolar bosons in two-dimensional optical lattices. Our study is motivated by the recent experimental realization of binary dipolar condensate mixtures of Er-Dy [Phys. Rev. Lett. 121, 213601 (2018)]. We model the system by using the extended two-species Bose-Hubbard model and calculate the ground-state phase diagrams by using mean-field theory. For selected cases we also obtain analytical phase boundaries by using the site-decoupled mean-field theory. For comparison we also examine the phase diagram of two-species Bose-Hubbard model. Our results show that the quantum phases with the long-range intraspecies interaction phase separate with no phase ordering. The introduction of the long-range interspecies interaction modifies the quantum phases of the system. It leads to the emergence of phase-separated quantum phases with phase ordering. The transition from the phase-separated quantum phases without phase ordering to phase ordered ones breaks the inversion symmetry. △ Less

Submitted 9 October, 2020; v1 submitted 14 April, 2020; originally announced April 2020.

Comments: 16 pages, 6 figures, Published in the October issue of Phys. Rev. A

Journal ref: Phys. Rev. A 102, 043309 (2020)

Testing Unsatisfiability of Constraint Satisfaction Problems via Tensor Products

Authors: Daya Gaur, Muhammad Khan

Abstract: We study the design of stochastic local search methods to prove unsatisfiability of a constraint satisfaction problem (CSP). For a binary CSP, such methods have been designed using the microstructure of the CSP. Here, we develop a method to decompose the microstructure into graph tensors. We show how to use the tensor decomposition to compute a proof of unsatisfiability efficiently and in parallel… ▽ More We study the design of stochastic local search methods to prove unsatisfiability of a constraint satisfaction problem (CSP). For a binary CSP, such methods have been designed using the microstructure of the CSP. Here, we develop a method to decompose the microstructure into graph tensors. We show how to use the tensor decomposition to compute a proof of unsatisfiability efficiently and in parallel. We also offer substantial empirical evidence that our approach improves the praxis. For instance, one decomposition yields proofs of unsatisfiability in half the time without sacrificing the quality. Another decomposition is twenty times faster and effective three-tenths of the times compared to the prior method. Our method is applicable to arbitrary CSPs using the well known dual and hidden variable transformations from an arbitrary CSP to a binary CSP. △ Less

Submitted 31 January, 2020; originally announced February 2020.

Comments: ISAIM 2020, International Symposium on Artificial Intelligence and Mathematics

MSC Class: 68T20 ACM Class: I.2.8