arXiv subjects

Vasily Bokov

Publications and source records attributed to Vasily Bokov.

3 recordsLinked to original sources

arXiv2026

Quantifying the Value of Privileged Information Using a PAC-Bayesian Approach

In practice, various learning scenarios provide access to auxiliary features exclusively during training. Incorporating such data to enhance model performance gave rise to a paradigm known as Learning Using Privileged Information (LUPI). While this extra information is intended to improve the resulting model, establishing a generalized, cohesive understanding of how privileged information (PI) transfers useful knowledge remains a challenge. Vapnik's original theory and subsequent works offer performance guarantees in certain cases, but these results are inherently per-algorithm and rely on setting-specific proof approaches. Consequently, a more general framework explaining how and when PI transfers useful knowledge is still missing. To bridge this gap, we introduce an algorithm-agnostic, information-theoretic approach based on the PAC-Bayes framework. Rather than asking whether a particular algorithm exploits PI, we ask how much value it could offer: comparing the tightest achievable risk bound with and without PI yields its potential - an upper limit on the extractable gain. We introduce a metric that quantifies this potential directly from empirical training risk, bypassing the need for test-time data access, and validate our findings in both supervised and unsupervised settings. The results demonstrate a robust correspondence between our training-time metric and true test-time performance gains. Ultimately, this work takes a necessary step toward an information-theoretic understanding of LUPI, and quantifying the potential of privileged features before committing to a model.

cs.LG↗

arXiv2026

Machine learning with minimal use of quantum computers: Provable advantages in Learning Under Quantum Privileged Information (LUQPI)

Quantum machine learning (QML) is often listed as a promising candidate for useful applications of quantum computers, in part due to numerous proofs of possible quantum advantages. A central question is how small a role quantum computers can play while still enabling provable learning advantages over classical methods. We study an especially restricted setting in which a quantum computer is used only as a feature extractor: it acts independently on individual data points, without access to labels or global dataset information, is available only to augment the training set, and is not available at deployment. Training and deployment are therefore carried out by fully classical learners on a dataset augmented with quantum-generated features. We formalize this model by adapting the classical framework of Learning Under Privileged Information (LUPI) to the quantum case, which we call Learning Under Quantum Privileged Information (LUQPI). Within this framework, we show that even such minimally involved quantum feature extraction, available only during training, can yield exponential quantum-classical separations for suitable concept classes and data distributions under reasonable computational assumptions. We further situate LUQPI within a taxonomy of related quantum and classical learning settings and show how standard classical machinery, most notably the SVM+ algorithm, can exploit quantum-augmented data. Finally, we present numerical experiments in a physically motivated many-body setting, where privileged quantum features are expectation values of observables on ground states, and observe consistent performance gains for LUQPI-style models over strong classical baselines.

quant-ph↗

arXiv2021

Vertex and Energy Reconstruction in JUNO with Machine Learning Methods

The Jiangmen Underground Neutrino Observatory (JUNO) is an experiment designed to study neutrino oscillations. Determination of neutrino mass ordering and precise measurement of neutrino oscillation parameters $\sin^2 2θ_{12}$, $Δm^2_{21}$ and $Δm^2_{32}$ are the main goals of the experiment. A rich physical program beyond the oscillation analysis is also foreseen. The ability to accurately reconstruct particle interaction events in JUNO is of great importance for the success of the experiment. In this work we present a few machine learning approaches applied to the vertex and the energy reconstruction. Multiple models and architectures were compared and studied, including Boosted Decision Trees (BDT), Deep Neural Networks (DNN), a few kinds of Convolution Neural Networks (CNN), based on ResNet and VGG, and a Graph Neural Network based on DeepSphere. Based on a study, carried out using the dataset, generated by the official JUNO software, we demonstrate that machine learning approaches achieve the necessary level of accuracy for reaching the physical goals of JUNO: $σ_E=3\%$ at $E_\text{vis}=1~\text{MeV}$ for the energy and $σ_{x,y,z}=10~\text{cm}$ at $E_\text{vis}=1~\text{MeV}$ for the position.

physics.ins-det↗