skip to main content


Title: Improved neural network Monte Carlo simulation
The algorithm for Monte Carlo simulation of parton-level events basedon an Artificial Neural Network (ANN) proposed in Ref.~ is used toperform a simulation of H\to 4\ell H → 4 ℓ decay. Improvements in the training algorithm have been implemented toavoid numerical instabilities. The integrated decay width evaluated bythe ANN is within 0.7% of the true value and unweighting efficiency of26% is reached. While the ANN is not automatically bijective betweeninput and output spaces, which can lead to issues with simulationquality, we argue that the training procedure naturally prefersbijective maps, and demonstrate that the trained ANN is bijective to avery good approximation.  more » « less
Award ID(s):
2014071
NSF-PAR ID:
10284157
Author(s) / Creator(s):
; ;
Date Published:
Journal Name:
SciPost Physics
Volume:
10
Issue:
1
ISSN:
2542-4653
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. A large body of work has demonstrated that parameterized artificial neural networks (ANNs) can efficiently describe ground states of numerous interesting quantum many-body Hamiltonians. However, the standard variational algorithms used to update or train the ANN parameters can get trapped in local minima, especially for frustrated systems and even if the representation is sufficiently expressive. We propose a parallel tempering method that facilitates escape from such local minima. This methods involves training multiple ANNs independently, with each simulation governed by a Hamiltonian with a different driver strength, in analogy to quantum parallel tempering, and it incorporates an update step into the training that allows for the exchange of neighboring ANN configurations. We study instances from two classes of Hamiltonians to demonstrate the utility of our approach using Restricted Boltzmann Machines as our parameterized ANN. The first instance is based on a permutation-invariant Hamiltonian whose landscape stymies the standard training algorithm by drawing it increasingly to a false local minimum. The second instance is four hydrogen atoms arranged in a rectangle, which is an instance of the second quantized electronic structure Hamiltonian discretized using Gaussian basis functions. We study this problem in a minimal basis set, which exhibits false minima that can trap the standard variational algorithm despite the problem’s small size. We show that augmenting the training with quantum parallel tempering becomes useful to finding good approximations to the ground states of these problem instances.

     
    more » « less
  2. In this study, pressure drop ( ) across air-cooled heat sinks (HSs) are predicted using an artificial neural network (ANN). A multilayer feed-forward ANN architecture with two hidden layers is developed. Backpropagation algorithm is used for training the network, and the accuracy of the network is evaluated by the root mean square error. The input data for training the neural network is prepared through three-dimensional simulation of air inside the channels of heat sinks using a computational fluid dynamics (CFD) approach. The developed ANN-based model in this study predicts with a high accuracy and within of the CFD-based data. The present study suggests that developing an ANN-based model with a high level of accuracy overcomes the limitations of physics-based correlations that their accuracy strongly depends on identifying and implementing key variables that affect the physics of a thermo-fluid phenomenon. 
    more » « less
  3. Antonis Achilleos ; Dario Della Monica (Ed.)
    In this paper, we explore the descriptive complexity theory of finite groups by examining the power of the second Ehrenfeucht-Fraisse bijective pebble game in Hella's (Ann. Pure Appl. Log., 1989) hierarchy. This is a Spoiler-Duplicator game in which Spoiler can place up to two pebbles each round. While it trivially solves graph isomorphism, it may be nontrivial for finite groups, and other ternary relational structures. We first provide a novel generalization of Weisfeiler-Leman (WL) coloring, which we call 2-ary WL. We then show that the 2-ary WL is equivalent to the second Ehrenfeucht-Fraisse bijective pebble game in Hella's hierarchy. Our main result is that, in the pebble game characterization, only O(1) pebbles and O(1) rounds are sufficient to identify all groups without Abelian normal subgroups (a class of groups for which isomorphism testing is known to be in P; Babai, Codenotti, & Qiao, ICALP 2012). In particular, we show that within the first few rounds, Spoiler can force Duplicator to select an isomorphism between two such groups at each subsequent round. By Hella's results (ibid.), this is equivalent to saying that these groups are identified by formulas in first-order logic with generalized 2-ary quantifiers, using only O(1) variables and O(1) quantifier depth. 
    more » « less
  4. null (Ed.)
    Brain-inspired cognitive computing has so far followed two major approaches - one uses multi-layered artificial neural networks (ANNs) to perform pattern-recognition-related tasks, whereas the other uses spiking neural networks (SNNs) to emulate biological neurons in an attempt to be as efficient and fault-tolerant as the brain. While there has been considerable progress in the former area due to a combination of effective training algorithms and acceleration platforms, the latter is still in its infancy due to the lack of both. SNNs have a distinct advantage over their ANN counterparts in that they are capable of operating in an event-driven manner, thus consuming very low power. Several recent efforts have proposed various SNN hardware design alternatives, however, these designs still incur considerable energy overheads.In this context, this paper proposes a comprehensive design spanning across the device, circuit, architecture and algorithm levels to build an ultra low-power architecture for SNN and ANN inference. For this, we use spintronics-based magnetic tunnel junction (MTJ) devices that have been shown to function as both neuro-synaptic crossbars as well as thresholding neurons and can operate at ultra low voltage and current levels. Using this MTJ-based neuron model and synaptic connections, we design a low power chip that has the flexibility to be deployed for inference of SNNs, ANNs as well as a combination of SNN-ANN hybrid networks - a distinct advantage compared to prior works. We demonstrate the competitive performance and energy efficiency of the SNNs as well as hybrid models on a suite of workloads. Our evaluations show that the proposed design, NEBULA, is up to 7.9× more energy efficient than a state-of-the-art design, ISAAC, in the ANN mode. In the SNN mode, our design is about 45× more energy-efficient than a contemporary SNN architecture, INXS. Power comparison between NEBULA ANN and SNN modes indicates that the latter is at least 6.25× more power-efficient for the observed benchmarks. 
    more » « less
  5. We propose a sequential algorithm for learning sparse radial basis approximations for streaming data. The initial phase of the algorithm formulates the RBF training as a convex optimization problem with an objective function on the expansion weights while the data fitting problem imposed only as an ℓ∞-norm constraint. Each new data point observed is tested for feasibility, i.e., whether the data fitting constraint is satisfied. If so, that point is discarded and no model update is required. If it is infeasible, a new basic variable is added to the linear program. The result is a primal infeasible-dual feasible solution. The dual simplex algorithm is applied to determine a new optimal solution. A large fraction of the streaming data points does not require updates to the RBF model since they are similar enough to previously observed data and satisfy the data fitting constraints. The structure of the simplex algorithm makes the update to the solution particularly efficient given the inverse of the new basis matrix is easily computed from the old inverse. The second phase of the algorithm involves a non-convex refinement of the convex problem. Given the sparse nature of the LP solution, the computational expense of the non-convex algorithm is greatly reduced. We have also found that a small subset of the training data that includes the novel data identified by the algorithm can be used to train the non-convex optimization problem with substantial computation savings and comparable errors on the test data. We illustrate the method on the Mackey-Glass chaotic time-series, the monthly sunspot data, and a Fort Collins, Colorado weather data set. In each case we compare the results to artificial neural networks (ANN) and standard skew-RBFs. 
    more » « less