Open Source
Gaussian Quasi-Likelihood Estimation for Nonsynchronously Observed L\'evy processes
arXiv stat.TH0 views
arXiv:2611.25329v1 Announce Type: new Abstract: We study parametric inference for a two-dimensional simple L\'evy process observed at high frequency over an expanding time horizon, where the two components are sampled in a possibly non-synchronous manner. Combining Gaussian quasi-likelihood estimation for L\'evy SDEs with the non-synchronous quasi-likelihood methodology, we construct a Gaussian-type quasi-likelihood based on the covariance structure of non-synchronous increments, and we prove consistency and asymptotic normality of the resulting estimators. The convergence rates coincide with those in the synchronous setting, while the asymptotic variance depends essentially on the sampling design and becomes more complicated under non-synchronicity. To handle the different moment structure of pure-jump L\'evy processes, we introduce an additional convergence assumption and provide tractable sufficient conditions linked to the observation scheme. Numerical experiments illustrate the finite-sample performance of the proposed estimators.
Source:https://arxiv.org/abs/2611.25329PreviousKernel Regression using Persistence DiagramsNextInference in High-Dimensional Linear Mediation Models
Related posts
Omni-modal decomposition autoencoders learn full-stack wearable disentangled representations
0arXiv eess.SP
Spectral Vision Transformer for Efficient Tokenization with Limited Data
0arXiv eess.SP
Triple Overlap in Pipelined Krylov Methods on Multi-GPU Systems
0arXiv cs.DC
Rhythm-TTS: High-Fidelity Shortcut Latent Modeling and Probabilistic Rhythm Generation for Zero-Shot Text-to-Speech
0arXiv cs.SD
Learning Integer Winding Evolution Times in Quantum Phase Estimation
0arXiv math.IT
Thirteen Improved Binary Constant-Weight Codes with Explicit Cyclic Orbit Certificates
0arXiv math.IT