Searcharxiv⌕ Search

arXiv subjects

Moshu Xu

Publications and source records attributed to Moshu Xu.

3 recordsLinked to original sources

Distributional change point analysis: high dimensionality, nonstationary dependence, and computational constraints

We study general distributional change point analysis for high-dimensional time series under nonstationary temporal dependence and limited computational resources. To mitigate noise accumulation from irrelevant coordinates and reduce the computational cost, we propose a sparse random projection procedure based on characteristic functions. We quantify the computational complexity of the random projections and show that the required number of projections grows polynomially with the ambient dimension. To accommodate nonstationary temporal dependence, a Gaussian multiplier bootstrap based on contrasts of local empirical characteristic functions is developed. We establish asymptotic false detection control and simultaneous guarantees for estimating the number and locations of multiple change points, with a near-parametric uniform localization rate up to logarithmic factors. A key technical innovation is a Gaussian approximation for high-dimensional multi-scale sparse time series, which is of independent interest. Simulation studies demonstrate the effectiveness of the proposed procedure in detecting and localizing distributional changes under various temporal dependence structures. An application to neural recordings further illustrates the practical utility of the proposed method in detecting diverse distributional changes in neural activity.

stat.ME↗

Probabilistic Limit Order Book Forecasting with Conditional Diffusion Models

Modeling limit order book (LOB) dynamics is central to market forecasting and execution-strategy evaluation. While diffusion models achieve strong empirical performance, existing LOB snapshot models typically lack a learned recurrent encoder and formal statistical guarantees. We introduce LOBGrad, a conditional DDPM framework for probabilistic LOB forecasting that combines an autoregressive RNN encoder with a diffusion-based generator. In particular, we establish an excess-risk oracle inequality for the one-step DDPM noise-prediction objective under exponential mixing and regularity assumptions. We further show that, under contractive dynamics, the dependence on the conditioning-window length can be reduced to a dependence on the latent-state dimension. Experiments on BTC/USDT, GOOGL, and META show that LOBGrad improves on its encoder-free diffusion ablation across all reported queue metrics and obtains competitive results against other baselines. It also reproduces the key stylized facts of LOB microstructure, and yields reasonably calibrated queue-based predictive intervals. A stylized downstream task illustrates the potential value of the generated forecasts.

cs.LG↗

Multicollinearity-agnostic feature screening for non-Euclidean responses: a factor adjusted approach

In high-dimensional settings, multicollinearity is a pervasive issue that can substantially impair the performance of feature screening methods based on marginal Fréchet regression. Feature screening for non-Euclidean responses becomes unreliable when ultrahigh-dimensional predictors suffer from multicollinearity, because feature-specific signals may be masked by shared latent factors. To mitigate this effect, we propose a Factor adjusted Fréchet sure independence screening procedure. The method first recovers latent common factors from the predictors and then evaluates each feature by the incremental Fréchet coefficient of determination contributed by its idiosyncratic component beyond the common factors. Under regularity conditions, we establish uniform approximation rates for the feasible screening utilities and prove the sure screening and sure ranking properties. Extensive numerical experiments provide compelling empirical support for the validity and effectiveness of our approach, particularly in scenarios with highly correlated covariates. We further illustrate the practical performance of our method through two representative non-Euclidean datasets: the ADNI dataset and the mortality dataset, both with distribution-valued responses.

stat.ME↗