Vectorizing Classical Tamil: Representation Learning for Verse-Commentary Pairs
研究通过构建古典泰米尔语诗-注释对语料库，使用多种深度学习模型探索信息表示学习，以解决文本内容和结构的自动理解与生成问题。
0 citationsRead paper
Selected work
Representative Papers
Quantum Annealing Enhanced Reinforcement Learning for Accurate Remaining Useful Lifetime Prediction
This study addresses the challenge of suboptimal convergence in remaining useful life prediction within high-dimensional non-convex spaces by proposing a novel integration of quantum annealing into a Q-learning framework. Specifically, each Q-value update is formulated as a Quadratic Unconstrained Binary Optimization (QUBO) problem and solved using the D-Wave Advantage quantum annealing system. The near-optimal action distributions generated through quantum annealing enhance exploration and effectively mitigate premature convergence. Combining minor embedding with a stochastic action selection mechanism, the proposed method achieves statistically significant improvements across six error metrics on both the NASA C-MAPSS dataset and a real-world equipment fleet maintenance dataset, outperforming classical and existing quantum baselines. These results demonstrate the practical viability and transformative potential of quantum annealing in predictive maintenance applications.
0 citationsRead paper
Petrov-Galerkin Variational Physics-Informed Neural Network Framework for Two-Dimensional Singularly Perturbed Problems
This work proposes a physics-informed neural network method based on a Petrov–Galerkin variational framework for two-dimensional singularly perturbed problems involving one or two small perturbation parameters. It is the first to integrate this variational framework with neural networks to efficiently resolve sharp boundary layers and multiscale features. The approach constructs a neural network trial solution space, employs tensor-product hat functions as test functions, computes source terms via automatic differentiation, and enforces Dirichlet boundary conditions strongly. Numerical experiments on benchmark problems demonstrate high accuracy in both the maximum norm and the $L^2$ norm, confirming the method’s effectiveness, robustness, and precise resolution of boundary layers in multiscale modeling.
0 citationsRead paper
MA-SBI: Misspecification-Aware Simulation-Based Inference via Side-Channel Guidance
This work addresses model misspecification in simulation-based inference arising from simplifying assumptions, particularly when ground-truth parameter–observation pairs are unavailable. It introduces the first calibration-free posterior correction framework that leverages unstructured side-channel information—such as textual labels or policy announcements—to learn a mapping from such auxiliary signals to shifts in observation space. Without retraining the simulator or requiring access to true parameters, the method enhances inference robustness. Theoretically, the correction efficacy is bounded by the mutual information between model misspecification and the side channel. Empirically, on a hidden benchmark, textual side information alone yields a posterior statistically equivalent to the oracle (validated via TOST), outperforming RoPE despite using less data. The approach also significantly improves predictive log-likelihood on real-world COVID/OxCGRT data while preserving posterior consistency in well-specified cognitive science tasks.
0 citationsRead paper
RAID: Semantic Graph Diffusion for True Cold-Start and Cross-Lingual Forecasting
This work addresses the challenges of genuine cold-start time series forecasting—where no historical observations are available—and cross-lingual generalization by proposing a novel approach based on semantic retrieval and graph-conditioned diffusion. The method constructs an inductive retrieval graph in a shared semantic space using frozen multilingual embeddings, first aggregating semantic neighbors to produce an initial forecast and then refining residual uncertainty through a gated diffusion module. It is the first to integrate semantic graph diffusion into cold-start prediction, enabling zero-shot cross-lingual transfer and supporting efficient non-autoregressive inference. Under strict cold-start evaluation protocols, the model significantly outperforms existing baselines in both point forecasting accuracy and prediction interval coverage, while reducing inference latency by an order of magnitude.
0 citationsRead paper
Recent publications
Latest Papers
Vectorizing Classical Tamil: Representation Learning for Verse-Commentary Pairs
研究通过构建古典泰米尔语诗-注释对语料库，使用多种深度学习模型探索信息表示学习，以解决文本内容和结构的自动理解与生成问题。
0 citationsRead paper
Quantum Annealing Enhanced Reinforcement Learning for Accurate Remaining Useful Lifetime Prediction
This study addresses the challenge of suboptimal convergence in remaining useful life prediction within high-dimensional non-convex spaces by proposing a novel integration of quantum annealing into a Q-learning framework. Specifically, each Q-value update is formulated as a Quadratic Unconstrained Binary Optimization (QUBO) problem and solved using the D-Wave Advantage quantum annealing system. The near-optimal action distributions generated through quantum annealing enhance exploration and effectively mitigate premature convergence. Combining minor embedding with a stochastic action selection mechanism, the proposed method achieves statistically significant improvements across six error metrics on both the NASA C-MAPSS dataset and a real-world equipment fleet maintenance dataset, outperforming classical and existing quantum baselines. These results demonstrate the practical viability and transformative potential of quantum annealing in predictive maintenance applications.
0 citationsRead paper
Petrov-Galerkin Variational Physics-Informed Neural Network Framework for Two-Dimensional Singularly Perturbed Problems
This work proposes a physics-informed neural network method based on a Petrov–Galerkin variational framework for two-dimensional singularly perturbed problems involving one or two small perturbation parameters. It is the first to integrate this variational framework with neural networks to efficiently resolve sharp boundary layers and multiscale features. The approach constructs a neural network trial solution space, employs tensor-product hat functions as test functions, computes source terms via automatic differentiation, and enforces Dirichlet boundary conditions strongly. Numerical experiments on benchmark problems demonstrate high accuracy in both the maximum norm and the $L^2$ norm, confirming the method’s effectiveness, robustness, and precise resolution of boundary layers in multiscale modeling.
0 citationsRead paper
MA-SBI: Misspecification-Aware Simulation-Based Inference via Side-Channel Guidance
This work addresses model misspecification in simulation-based inference arising from simplifying assumptions, particularly when ground-truth parameter–observation pairs are unavailable. It introduces the first calibration-free posterior correction framework that leverages unstructured side-channel information—such as textual labels or policy announcements—to learn a mapping from such auxiliary signals to shifts in observation space. Without retraining the simulator or requiring access to true parameters, the method enhances inference robustness. Theoretically, the correction efficacy is bounded by the mutual information between model misspecification and the side channel. Empirically, on a hidden benchmark, textual side information alone yields a posterior statistically equivalent to the oracle (validated via TOST), outperforming RoPE despite using less data. The approach also significantly improves predictive log-likelihood on real-world COVID/OxCGRT data while preserving posterior consistency in well-specified cognitive science tasks.
0 citationsRead paper
RAID: Semantic Graph Diffusion for True Cold-Start and Cross-Lingual Forecasting
This work addresses the challenges of genuine cold-start time series forecasting—where no historical observations are available—and cross-lingual generalization by proposing a novel approach based on semantic retrieval and graph-conditioned diffusion. The method constructs an inductive retrieval graph in a shared semantic space using frozen multilingual embeddings, first aggregating semantic neighbors to produce an initial forecast and then refining residual uncertainty through a gated diffusion module. It is the first to integrate semantic graph diffusion into cold-start prediction, enabling zero-shot cross-lingual transfer and supporting efficient non-autoregressive inference. Under strict cold-start evaluation protocols, the model significantly outperforms existing baselines in both point forecasting accuracy and prediction interval coverage, while reducing inference latency by an order of magnitude.
0 citationsRead paper