Research
On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

168,695 papers · 148 categories

Trend · papers per month

35810 · May 202619922001200920172026
48 results for RMSE

Paper presents a novel approach to predict volatility using robust least squares method.

problem Challenges in predicting volatility due to irregularities, high fluctuations, and noise in financial time series.
method Robust least squares method applied in two approaches: with and without least absolute residuals (LAR).
result Robust least squares method with LAR approach yields better results for volatility and its components.

This research predicts cryptocurrency price volatility using deep learning models.

problem Predicting the volatility of cryptocurrency prices to mitigate investment risk.
method Used CNN, LSTM, BiLSTM, and GRU models to predict the risk factor of 20 cryptocurrency parameters.
result Developed a new model with RMSE of 0.0089, significantly outperforming existing models.

A new algorithm for missing data imputation with low RMSE and explainability.

problem Missing data in various domains, especially in critical applications requiring low RMSE and explainability.
method DIMV algorithm that uses conditional distribution of features based on fully observed features.
result DIMV provides low RMSE, scalability, and explainability for imputed values.

Informer model with GMADL loss outperforms benchmarks in high frequency Bitcoin trading.

problem Developing automated trading strategies for high frequency Bitcoin data.
method Informer architecture with RMSE, GMADL, and Quantile loss functions.
result Informer model with GMADL loss function outperforms benchmarks in trading outcomes.

The paper evaluates company investment value using machine learning models.

problem Evaluating the investment value of companies based on machine learning.
method Data mining, feature selection, cross-validation, stacking model, Bayesian Ridge Regression.
result The RMSE of the final model is 3.047, indicating improved stability and generalization.

Study develops ensemble machine learning framework for predicting groundwater heavy metal pollution.

problem Statistical complexity and spatial heterogeneity of heavy metal contamination in groundwater.
method Nested cross-validated ensemble machine learning with response transformations (raw, log, Gaussian copula).
result Copula-based models with DBSCAN clustering diagnostics provide the most reliable and interpretable assessments of groundwater contamination.

Bayesian Neural Networks improve geophysical model ensembles with reduced uncertainty.

problem Improving geophysical model projections and uncertainty quantification.
method Developed a Bayesian Neural Network ensemble strategy for geophysical models.
result Bayesian Neural Network ensemble outperforms existing methods in ozone prediction.

Proposes a framework to handle missing data in traffic forecasting with sensor blackouts.

problem Missing data in traffic forecasting due to sensor blackouts, especially when correlated with traffic conditions.
method Latent state-space framework that models traffic dynamics and sensor dropout probabilities.
result Improves traffic forecasting by reducing blackout imputation RMSE from 7.02 to 4.23, with MNAR modeling providing additional gains.

Generative model improves noise estimation in stochastic rotating shallow water models.

problem Improving noise estimation in stochastic partial differential equations for fluid dynamics.
method Replaced PCA with a generative model to avoid constraints on stochastic increments.
result Generative model produces better RMSE, CRPS score, and forecast rank histograms.

A new Bayesian model improves forecasting for intermittent demand.

problem Sparse observations, cold-start items, and obsolescence in intermittent demand forecasting.
method Hierarchical Bayesian TSB model with partial pooling and calibrated probabilistic configuration.
result TSB-HB achieves the lowest RMSE and RMSSE on the UCI Online Retail dataset.

Novel approach uses ENN for UQ in gust predictions, reducing RMSE and improving confidence.

problem Reducing bias and uncertainty in wind gust predictions.
method Evidential Neural Network (ENN) with Explainable AI.
result 47% reduction in RMSE, 95% coverage of observed gusts at 179 out of 266 stations.

SHIFT improves robustness in estimating dose-response functions with heavy-tailed contamination.

problem Outliers bias estimates of average dose-response functions in heavy-tailed data.
method SHIFT combines cross-fit nuisance orthogonalization, Welsch-loss, and defensive OLS refit.
result SHIFT reduces RMSE from 1.03 to 0.33 on localized contamination test.

This paper evaluates forecast quality in electricity markets beyond traditional accuracy measures.

problem Traditional accuracy measures fail to reflect the economic value of electricity price forecasts.
method Investigates four quality dimensions: accuracy, dispersion, association, and extremum identification.
result Dispersion- and association-based measures better capture forecast economic value.

This paper uses Bayesian optimization to efficiently identify stochastic dynamical systems.

problem Efficiently identifying linear stochastic dynamical systems with unknown coefficients and noise variances.
method Adaptive Bayesian optimization with ensemble Gaussian processes (EGP) and Kalman filter recursion.
result BO-based estimator achieves RMSE below the Cramer-Rao bound, improving robustness and consistency.

MELO predicts electricity loads by adapting to shifts without external indicators.

problem Adapting to non-stationary prediction challenges in online settings.
method MELO combines multiple forgetting factors and aggregation rules to adaptively predict.
result MELO reduces RMSE by 34.7% compared to base predictors and external covariates.

Study improves forex forecasting accuracy using machine learning models.

problem Improving accuracy in predicting foreign exchange rates.
method Employed LSTM neural networks and Gradient Boosting Classifier for forecasting.
result Achieved 99.449% accuracy in forecasting USD/BDT exchange rates.

A new sparse benchmark metabench identifies key abilities from large benchmarks.

problem Redundancy and compression in existing benchmarks.
method Data from 5000+ LLMs to identify most informative items, distilling a sparse benchmark.
result Sparse benchmark metabench captures underlying abilities with high accuracy.

Crop yield is a highly complex trait determined by multiple factors such as genotype, environment, and their interactions. Accurate yield prediction requires fundamental understanding of the functional relationship between yield and these interactive factors, and to reveal such relationship requires both comprehensive …

2019-02-07abs ↗pdf ↗

Markov chain Monte Carlo (MCMC) algorithms are ubiquitous in Bayesian computations. However, they need to access the full data set in order to evaluate the posterior density at every step of the algorithm. This results in a great computational burden in big data applications. In contrast to MCMC methods, Stochastic Gra…

2016-09-15abs ↗pdf ↗

Study on Neural Tangent Kernel of Matrix Product States and their convergence.

problem Understanding the convergence of Neural Tangent Kernel of Matrix Product States.
method Analyzing the Neural Tangent Kernel of Matrix Product States and proving its convergence in the infinite bond dimensional limit.
result The Neural Tangent Kernel of Matrix Product States converges to a constant matrix during training.

Paper proposes combining GAM and DNN for accurate peak demand estimation from lower-resolution data.

problem Predicting high-resolution peak demand from limited lower-resolution data.
method Combines generalized additive models (GAM) and deep neural networks (DNN) for half-hourly load forecasting.
result Proposed method reduces out-of-sample RMSE by 57.4% compared to benchmark.

The paper predicts Bitcoin prices using machine learning and sentiment analysis.

problem Predicting the future price of Bitcoin in USD.
method Applied supervised machine learning and sentiment analysis to Twitter and Reddit data.
result LSTM models with multi-feature analysis outperformed ARIMA models in predicting Bitcoin prices.

New method uses DTW to evaluate neural network forecasts of geomagnetic indices.

problem Evaluation metrics fail to capture persistence behavior in neural network forecasts.
method Dynamic Time Warping (DTW) to measure time series similarity, training neural networks to remove persistence.
result DTW reveals persistence behavior in neural network forecasts, confirming visual inspection.

A neural network solves Black-Scholes PDE for option pricing with uncertainty quantification.

problem Solving the Black-Scholes equation for option pricing with uncertainty.
method Physics-informed neural network (PINN) that embeds BS operator and conditions, handles early exercise via relaxation, and uses anchored-ensemble fine-tuning for uncertainty quantification.
result The method achieves low errors and accurate predictions for European and American options, outperforming data-driven baselines.