This work improves Recursive Neural Gas (RNG) for reservoir computing.
problem Improving performance of fully-trainable reservoirs in Recurrent Neural Networks (RNN).
method Describes an accurate model of RNG and shows comparative results on three datasets.
result RNG-based reservoirs can achieve better performance under specific circumstances.
In this paper we treat a gas storage valuation problem as a Markov Decision Process. As opposed to existing literature we model the gas price process as a regime-switching model. Such a model has shown to fit market data quite well in Chen and Forsyth (2010). Before we apply a numerical algorithm to solve the problem, …
Paper improves gas species identification in complex mixtures using neural networks.
problem Identifying gas species in multi-gas mixtures with high accuracy.
method Multi-label neural networks with optimal thresholding for IR spectroscopy.
result Optimal thresholding improves classification performance over conventional methods.
Paper uses neural networks to predict NOx emissions from gas turbines.
problem Predicting NOx emissions from degrading gas turbines.
method Applied neural network algorithm to model NOx emissions from nine process variables.
result Neural network model optimizes process variables for minimal NOx emissions.
GA improves deep neural networks by creating sparser models.
problem Training deep neural networks efficiently and sparsely.
method Genetic algorithms for evolving deep neural networks.
result GA-assisted deep autoencoder produces sparser neural networks.
This study compares GNNs and GA-MLPs, finding GA-MLPs can distinguish graphs but not count walks.
problem Comparing expressive power and graph isomorphism testing capabilities of GNNs and GA-MLPs.
method GA-MLPs augment node features with multi-hop operators and apply MLPs node-wise; GNNs are compared as a baseline.
result GA-MLPs can distinguish almost all non-isomorphic graphs but cannot count attributed walks, unlike GNNs.
Adds recursion to deep learning frameworks for better handling of recursive data structures.
problem Lack of support for recursion in existing deep learning frameworks.
method Complements existing frameworks with recursive execution of dataflow graphs and APIs for recursive definitions.
result Recursive implementation reduces training and inference time by more effectively using resources.
GAS-Norm improves deep learning time series forecasting in non-stationary settings.
problem Deep learning models struggle with non-stationary time series data.
method Combines GAS model for adaptive normalization with deep neural networks.
result Improves deep learning performance in 21 out of 25 settings.
Improved gas demand forecasting using ensemble methods.
problem Short-term prediction of gas demand components.
method Nine base forecasters (Ridge Regression, GP, NN, ANN, Torus, LASSO, Elastic Net, RF, SVR) and four ensemble predictors (simple, weighted, subset, SVR aggregation) were evaluated.
result Ensemble predictors outperformed individual base forecasters and TSO predictions.
Deep learning improves combustor anomaly detection in gas turbines.
problem Improving anomaly detection performance in gas turbine combustors.
method Hierarchically learned features from exhaust gas temperature sensor measurements using deep learning.
result Deep learning-based anomaly detection significantly improved combustor anomaly detection performance.
The paper compares machine learning models for forecasting residential gas demand, highlighting the impact of temperature forecasts.
problem Forecasting residential gas demand for optimal energy planning.
method Implemented and compared five models: Ridge Regression, GP, k-Nearest Neighbour, ANN, and Torus Model.
result ANN is the best model in terms of RMSE, while GP is the best in terms of MAE.
Model predicts EMF of Ni-Mn-Ga MSMA, improved with GRNN.
problem Predicting the electromotive force (EMF) of Ni-Mn-Ga MSMA under various conditions.
method Developed a new constitutive model for Ni-Mn-Ga single crystals, incorporating magnetic easy axis offset. Used GRNN to enhance model predictions.
result GRNN improves model predictions of EMF, capturing more experimental features.
Paper tackles few-shot class-incremental learning with a neural gas network.
problem Incrementally learn new classes from very few labelled samples without forgetting old classes.
method Proposes TOPIC framework using a neural gas network to preserve class topology and adapt to new samples.
result Significantly outperforms other methods on CIFAR100, miniImageNet, and CUB200 datasets.
GA method reduces gradient staleness in cloud computing.
problem Gradient staleness in asynchronous SGD methods.
method Gap-Aware (GA) method that penalizes stale gradients linearly to the Gap.
result GA outperforms existing methods in final test accuracy.
Neural network for water treatment anomaly detection with GA architecture optimization.
problem Detect anomalies in water treatment systems.
method Genetic algorithms for NN architecture optimization, NAB metric, F1-metric drawbacks analysis, techniques to improve AD quality.
result Improved anomaly detection quality through genetic algorithms and techniques.
Bayesian neural networks improve uncertainty in data-driven VFMs for oil and gas wells.
problem Uncertainty and robustness in data-driven VFMs for oil and gas wells.
method Bayesian neural networks with variational inference for uncertainty quantification.
result Variational inference provides more robust predictions on future data.
Developed AI models for multi-gas detection in near IR spectrums.
problem Detecting multiple gases in near IR spectrums.
method Used Monte Carlo KNN and multi-resolution CNN, synthesized near IR spectrums, optimized kernel sizes and channels.
result Multi-resolution CNN outperforms other models.
This paper presents the R package GAS for the analysis of time series under the Generalized Autoregressive Score (GAS) framework of Creal et al. (2013) and Harvey (2013). The distinctive feature of the GAS approach is the use of the score function as the driver of time-variation in the parameters of nonlinear models. T…
Developed neural network for predicting mechanical properties of composite materials.
problem Predicting and optimizing mechanical properties of composite materials.
method Convolutional neural network model integrated with a genetic algorithm optimizer.
result Highly accurate predictions and optimal microstructural designs identified.
GAS models have been recently proposed in time-series econometrics as valuable tools for signal extraction and prediction. This paper details how financial risk managers can use GAS models for Value-at-Risk (VaR) prediction using the novel GAS package for R. Details and code snippets for prediction, comparison and back…
For a given real generic curve $\ga: S^1\to \Bbb {RP}^n$ let $D_\ga$ denote the ruled hypersurface in RPn consisting of all osculating subspaces to $\ga$ of codimension 2. A curve $\ga: S^1\to \Bbb {RP}^n$ is called convex if the total number of its intersection points (counted with multiplicities) with any h…
Recurrent neural networks (RNNs) process input text sequentially and model the conditional transition between word tokens. In contrast, the advantages of recursive networks include that they explicitly model the compositionality and the recursive structure of natural language. However, the current recursive architectur…
The paper uses tensor decompositions to improve neural network models for tree data.
problem Encoding structural knowledge from tree-structured data efficiently.
method Introduces new aggregation functions using Canonical and Tensor-Train decompositions.
result Proposed models outperform traditional methods on tree classification tasks.
Automated translation of mathematical formulae using recursive neural networks.
problem Performing translations between different representations of mathematical formulae.
method Recursive neural networks with multi-variate multi-valued Long Short-Term Memory cells and novel clustering and mini-batch training techniques.
result Achieved a prediction accuracy of 47.05% for predicting symbols at the correct position and 92.3% when ignoring the predicted position.
Paper generalizes GCNNs using edge-variant recursions for better graph signal classification.
problem Improving graph signal classification performance.
method Formulates a general framework for GCNNs using edge-variant graph filters.
result Shows superior performance in graph signal classification problems.
Modified deep LSTM model with EnKF improves gas flow rate predictions in mature gas wells.
problem Predicting gas production from mature gas wells with high accuracy and robustness.
method Modified deep LSTM model for flow rate prediction and EnKF for updating predictions.
result EnKF updated model leads to better flow predictions with lower Jeffreys' J-divergences.
New method reduces clustering time and improves accuracy.
problem High time and space complexity in spectral clustering.
method Approximate spectral clustering using GNG network topology.
result Equal or better clustering performance than traditional SC.
Algorithm discovers dynamic cell structures for better neural network performance.
problem Finding optimal neural network architectures for diverse data samples and time steps.
method Combines recurrent and recursive neural networks to dynamically search for customized cell structures.
result Achieves better prediction accuracy compared to existing models.
This paper analyzes Ethereum's gas fees and their derivatives, providing a comprehensive model.
problem Understanding and predicting gas fees on the Ethereum blockchain.
method Analyzed Ethereum's gas fee structure and used a fractional Ornstein-Uhlenbeck process to model gas prices.
result A model for pricing and trading gas fee derivatives to hedge against volatility.
Gradient GA uses gradient information to improve molecular design.
problem Random walk exploration limits genetic algorithms' quality and speed in molecular design.
method Gradient GA incorporates gradient information from the objective function into genetic algorithms, using a differentiable neural network and Discrete Langevin Proposal.
result Significantly improves convergence speed and solution quality over traditional genetic algorithms.
Geodesic X-ray transform proves injective for smooth one-forms on gas giant manifolds.
problem Injectivity of geodesic X-ray transform for one-forms on specific manifolds.
method Pestov identity and asymptotic analysis of short geodesics.
result Geodesic X-ray transform is solenoidally injective for smooth one-forms on gas giant manifolds.
Recursive KalmanNet combines neural networks with Kalman filters for precise state estimation.
problem State estimation in systems with noisy measurements and non-Gaussian noise.
method Recursive KalmanNet uses a recurrent neural network to estimate states with consistent error covariance, optimizing for Gaussian negative log-likelihood.
result Recursive KalmanNet outperforms conventional Kalman filters and deep learning-based estimators in non-Gaussian noise conditions.
Recursive neural networks improve quark/gluon tagging performance.
problem Improving quark/gluon tagging accuracy using machine learning.
method Recursive neural networks (RecNNs) that embed jet clustering history recursively.
result RecNNs outperform traditional boosted decision tree (BDT) by a few percent in gluon rejection rate.
Study the Hessian geometry of an ideal gas in a centrifuge.
problem Understanding the Hessian geometry of an ideal gas in a centrifuge.
method Investigate the Hessian geometry associated with an ideal gas in a spherical centrifuge, using the action of the Euclidean rotation group.
result The Hessian geometry of a spherical rigid body is isometric to a hyperbolic space in the high angular velocity limit.
Introduces GA-P/E, a growth-adjusted stock valuation measure.
problem Evaluating stock value and predicting future returns.
method Computes a payback period adjusted for earnings growth, using a sorted portfolio methodology.
result Low GA-P/E stocks outperform high GA-P/E stocks in absolute and risk-adjusted returns.
Improved GAS models using trees and forests for better forecasts.
problem Improving forecasts from GAS models to avoid curse of dimensionality.
method Localized parameters using decision trees and random forests.
result Significantly outperform baseline GAS model in empirical analyses.
Machine learning detects underwater gas leaks.
problem Early detection of gas leaks in underwater reservoirs.
method Machine learning and Passive Acoustic Monitoring (PAM).
result Classification algorithms achieve good performance in detecting leaks.
Paper proposes GAS-ALD model for financial risk prediction.
problem Skewed distribution of financial return data.
method Generalized autoregressive score (GAS) framework with asymmetric Laplace distribution (ALD).
result GAS-ALD model predicts VaR and ES more accurately than traditional models.
Support Vector Machines predict gas-liquid flow patterns with 97% accuracy.
problem Predicting gas-liquid flow patterns in multiphase flow systems.
method Support Vector Machine (SVM) applied to a dataset of two-phase flow patterns.
result Achieved 97% correct classification of flow patterns.
Optimizes routing in decentralized exchanges with gas fees.
problem Routing in decentralized exchanges with fixed gas fees.
method General optimization framework with mixed-integer model, incorporating gas fees.
result Explicit Karush-Kuhn-Tucker system linking prices, fees, and activation.
Study the geometry of gas giant planets to infer their internal structure.
problem Determine the interior structure of gas giant planets using boundary data.
method Geometric analysis of Riemannian manifolds with conformal blow-up at the boundary.
result The interior structure of a gas giant is uniquely determined by different types of boundary data.
We calculate the free energy of Coulomb gas systems on Riemann surfaces.
problem Analyzing the free energy of Coulomb gas systems on Riemann surfaces.
method Using bosonization formula and analytic torsion, we derive the asymptotic expansion of the partition function.
result We prove the geometric version of the Zabrodin-Wiegmann conjecture in the determinantal case.
New ODE solvers improve training efficiency and accuracy.
problem Training Neural ODEs requires efficient and accurate gradient calculation.
method Presented algebraically reversible ODE solvers that are time and memory efficient, calculate exact gradients, and are numerically stable.
result Reversible solvers strictly improve upon previous architectures in efficiency and accuracy.
A co-evolutionary approach for Heston model calibration reduces overfitting with diverse datasets.
problem Overfitting and lack of generalization in Heston model calibration.
method Coupling a genetic algorithm with an evolving neural inverse map, using both GA-history sampling and Latin hypercube sampling.
result Diverse datasets improve out-of-sample stability and calibration accuracy.
Modeling gas fee competition in decentralized exchanges to optimize arbitrage profits.
problem Gas fees and transaction ordering in decentralized exchanges create arbitrage opportunities.
method Developed a first equilibrium model of gas fee competition between two arbitrageurs under three transaction reversion settings.
result Mixed equilibria exist, and their characteristics depend on inventory risk and transaction settings.
Deep learning solves dynamic programming with recursive utility.
problem Challenges in solving high-dimensional discrete-time dynamic programming problems with recursive utility.
method Certainty Equivalent Learning (CEL) algorithm that learns certainty-equivalent value directly with neural networks.
result Accurate value and policy approximations in high-dimensional problems, comparable to VFI in some cases.
Graph neural networks improve molecular property prediction.
problem Efficiently predicting molecular properties with high accuracy and scalability.
method Gated Graph Recursive Neural Networks (GGNN) with skip connections.
result GGNN achieves state-of-the-art performance on molecular property prediction benchmarks.
The study finds a long-term relationship between Dubai crude oil and US natural gas prices.
problem Examining the relationship between Dubai crude oil and US natural gas prices.
method Used unit root and cointegration tests, ARDL cointegration technique, and Toda-Yamamoto causality test.
result There is a long-run relationship with unidirectional causality from Dubai crude oil to US natural gas.