arXiv research

Paper analyzes distributed learning with non-i.i.d. samples.

There has been considerable advance in understanding the properties of sparse regularization procedures in high-dimensional models. In time series context, it is mostly restricted to Gaussian autoregressions or mixing sequences. We study oracle properties of LASSO estimation of weakly sparse vector-autoregressive model…

Uniform counting formulas for orthogeodesics in Kleinian groups converge.

A new perspective on self-attention models using MLPs.

Research on mixed polynomials, extending non-degeneracy concepts to complex variables.

Two algorithms learn Gaussian graphical models from Glauber dynamics trajectories, achieving optimal performance.

The study explores splitting conditions for mixed braid group sequences.

Strong stability of ergodic iterations proven without ergodic driving sequence.

Traffic forecasting problem remains a challenging task in the intelligent transportation system due to its spatio-temporal complexity. Although temporal dependency has been well studied and discussed, spatial dependency is relatively less explored due to its large variations, especially in the urban environment. In thi…

New method for robust matrix completion with mixed data types.

Estimates missing mass in Markovian sequences with linear runtime and near-optimal risk.

The paper analyzes convergence rates of Langevin dynamics and Proximal Sampler using $Φ$-divergence.

Log-concave coefficient sequences for two-bridge knots proved.

Inference in general Ising models is difficult, due to high treewidth making tree-based algorithms intractable. Moreover, when interactions are strong, Gibbs sampling may take exponential time to converge to the stationary distribution. We present an algorithm to project Ising model parameters onto a parameter set that…

Dark Experience improves continual learning with a simple, strong baseline.

Paper analyzes Nyström regularization for time series forecasting with sequential sub-sampling.

Models for sequential data such as the recurrent neural network (RNN) often implicitly model a sequence as having a fixed time interval between observations and do not account for group-level effects when multiple sequences are observed. We propose a model for grouped sequential data based on the RNN that accounts for …

Estimates stationary mass and frequency from non-i.i.d. data.

Study online learning in RKHS with dependent processes, focusing on \(β\)- and \(φ\)-mixing.

Superpixel-mix enhances reliability in semantic segmentation.

Recent character and phoneme-based parametric TTS systems using deep learning have shown strong performance in natural speech generation. However, the choice between character or phoneme input can create serious limitations for practical deployment, as direct control of pronunciation is crucial in certain cases. We dem…

We provide a measure based topology for certain unions of C2 rectifiable submanifolds of mixed dimensions in Rn. In this topology lower dimensional sets remain in the limit as measures when higher dimensional sets collapse down to them. For example a decreasing sequence of spheres may have a limit consisting of just a …

Paper describes links of mixed polynomials with specific properties.

New method uses dendrograms for better mixture model selection and clustering.

Study differentially private methods for learning Hawkes processes.

The paper analyzes variational autoencoders for state space models with risk bounds.

Study continuity of limit sets in symmetric spaces.

Bayesian method learns causal orderings from heterogeneous data.

LES optimizes designs by sampling descent sequences, achieving strong sample efficiency.

Gibbs sampler contracts entropy under strong log-concavity, improving mixing time.

In this paper we present the construction of explicit quasi-isomorphisms that compute the cyclic homology and periodic cyclic homology of crossed-product algebras associated with (discrete) group actions. In the first part we deal with algebraic crossed-products associated with group actions on unital algebras over any…

SelectMix improves deep learning robustness against noisy labels.

MARGE learns to reconstruct text by paraphrasing, achieving strong performance across multiple tasks.

In this note, given a polarized algebraic manifold $(X,L)$, we define the Donaldson-Futaki invariant for a sequence of test configurations for $(X,L)$ with exponents tending to infinity. This then allows us to define a strong version of K-stability or K-semistability for $(X,L)$. In particular, $(X,L)$ will be shown to…

Paper extends nonparametric regression bounds for dependent $\beta$-mixing samples.

Graph matching in noisy environments with Markovian errors.

This paper proposes a new meta-learning method -- named HARMLESS (HAwkes Relational Meta LEarning method for Short Sequences) for learning heterogeneous point process models from short event sequence data along with a relational network. Specifically, we propose a hierarchical Bayesian mixture Hawkes process model, whi…

We study a special case of the problem of statistical learning without the i.i.d. assumption. Specifically, we suppose a learning method is presented with a sequence of data points, and required to make a prediction (e.g., a classification) for each one, and can then observe the loss incurred by this prediction. We go …

Study on multitask learning performance factors.

We show that the graph TQFT for Heegaard Floer homology satisfies a strong version of Atiyah's duality axiom for a TQFT. As an application, we compute some Heegaard Floer mixed invariants of 4-dimensional mapping tori in terms of Lefschetz numbers on $HF^+$.

Recently Swatee Naik and Theodore Stanford proved that two S-equivalent knots are related by a finite sequence of doubled-delta moves on their knot diagrams. We show that classical S-equivalence is not sufficient to extend their result to ordered links. We define a new algebraic relation on Seifert matrices, called Str…

We address the problem of estimating the mixing time of a Markov chain from a single trajectory of observations. Unlike most previous works which employed Hilbert space methods to estimate spectral gaps, we opt for an approach based on contraction with respect to total variation. Specifically, we estimate the contracti…

This work finds mixed equilibria in zero-sum games using interacting particle dynamics.

New method combines domain changes and sparse mixing for better latent variable learning.

Quantum mixing for eigenfunctions on hyperbolic surfaces converging to the hyperbolic plane.

A universal framework for constructing confidence sets using sequential likelihood mixing.

Not all types of supervision signals are created equal: Different types of feedback have different costs and effects on learning. We show how self-regulation strategies that decide when to ask for which kind of feedback from a teacher (or from oneself) can be cast as a learning-to-learn problem leading to improved cost…

Faster convergence of kernel mean embeddings using variance information.