Search Results for author:
Found 10 papers, 6 papers with code
Multivariate Online Linear Regression for Hierarchical Forecasting
In this paper, we consider a deterministic online linear regression model where we allow the responses to be multivariate.
Non-asymptotic Analysis of Biased Adaptive Stochastic Approximation
This paper provides a comprehensive non-asymptotic analysis of SGD with biased gradients and adaptive steps for convex and non-convex smooth functions.
Dynamical Survival Analysis with Controlled Latent States
We consider the task of learning individual-specific intensities of counting processes from a set of static variables and irregularly sampled time series.
New directions in the applications of rough path theory
This article provides a concise overview of some of the recent advances in the application of rough path theory to machine learning.
Learning the Dynamics of Sparsely Observed Interacting Systems
We address the problem of learning the dynamics of an unknown non-parametric system linking a target and a feature time series.
Scaling ResNets in the Large-depth Regime
initializations, the only non-trivial dynamics is for $\alpha_L = 1/\sqrt{L}$ (other choices lead either to explosion or to identity mapping).
Framing RNN as a kernel method: A neural ODE approach
Building on the interpretation of a recurrent neural network (RNN) as a continuous-time neural differential equation, we show, under appropriate conditions, that the solution of a RNN can be viewed as a linear function of a specific feature set of the input sequence, known as the signature.
Functional linear regression with truncated signatures
We place ourselves in a functional regression setting and propose a novel methodology for regressing a real output on vector-valued functional covariates.
Methodology 62R10 (Primary), 60L10 (Secondary)
A Generalised Signature Method for Multivariate Time Series Feature Extraction
There is a great deal of flexibility as to how this method can be applied.
Embedding and learning with signatures
It is shown that a specific embedding, called lead-lag, is systematically the strongest performer across all datasets and algorithms considered.
