Author name cluster

Kun Gao

Possible papers associated with this exact author name in Arrow. This page groups case-insensitive exact name matches and is not a full identity disambiguation profile.

5 papers

1 author row

AAAI Conference 2026 Conference Paper

R-Tuning: Wavelet-Decomposed Replay and Semantic Alignment for Continual Adaptation of Pretrained Time-Series Models

Tianyi Yin
Jingwei Wang
Chenze Wang
Han Wang
Jiexuan Cai
Min Liu
Yunlong Ma
Kun Gao

Pre-trained models have demonstrated exceptional generalization capabilities in time-series forecasting; however, adapting them to evolving data distributions remains a significant challenge. A key hurdle lies in accessing the original training data, as fine-tuning solely on new data often leads to catastrophic forgetting. To address this issue, we propose Replay Tuning (R-Tuning), a novel framework designed for the continual adaptation of pre-trained time-series models. R-Tuning constructs a unified latent space that captures both prior and current task knowledge through a frequency-aware replay strategy. Specifically, it augments model-generated samples via wavelet-based decomposition across multiple frequency bands, generating trend-preserving and fusion-enhanced variants to improve representation diversity and replay efficiency. To further reduce reliance on synthetic samples, R-Tuning introduces a latent consistency constraint that aligns new representations with the prior task space. This constraint guides joint optimization within a compact and semantically coherent latent space, ensuring robust knowledge retention and adaptation. Extensive experimental results demonstrate the superiority of R-Tuning, which reduces MAE and MSE by up to 46.9% and 46.8%, respectively, on new tasks, while preserving prior knowledge with gains of up to 5.7% and 6.0% on old tasks. Notably, under few-shot settings, R-Tuning outperforms all state-of-the-art baselines even when synthetic proxy samples account for only 5% of the new task dataset.

PDF Details DOI

EAAI Journal 2025 Journal Article

Context-aware inverse reinforcement learning for modeling individuals’ daily activity schedules

Dongjie Liu
Dawei Li
Kun Gao
Yuchen Song
Zijie Zhou

Understanding individual and crowd dynamics in urban environments is critical for numerous applications, such as urban planning, traffic forecasting, and location-based services. Therefore, accurately modeling individuals' daily activity schedules is essential. Traditional methods, like utility-based and rule-based approaches, rely on expert knowledge and presumed model structures. While machine learning methods offer flexibility, they often ignore explicit behavioral mechanisms, particularly comprehensive discussion and integration of context related to individuals' daily travel. To address these, we propose a framework that integrates travel context with deep Inverse Reinforcement Learning (IRL), learning temporal patterns from sociodemographics, start time and duration of the current activity, travel modes, and land use. Specifically, individuals' activity schedules are initially formulated as a Markov Decision Process to simulate travelers’ sequential decision-making processes, laying the groundwork for the IRL framework; Then, a context-aware IRL method is proposed to model individuals' travel decision-making from observed temporal trajectories. Finally, we validate the proposed model by demonstrating its superior performance over discrete choice model and several well-known imitation learning benchmarks in tasks such as policy performance comparison, reward recovery, model generalizability, and computational efficiency using travel behavior datasets. This approach provides meaningful behavioral insights and paves the way for Artificial Intelligence-driven activity schedulers modeling.

Details DOI

AIJ Journal 2024 Journal Article

A differentiable first-order rule learner for inductive logic programming

Kun Gao
Katsumi Inoue
Yongzhi Cao
Hanpin Wang

Learning first-order logic programs from relational facts yields intuitive insights into the data. Inductive logic programming (ILP) models are effective in learning first-order logic programs from observed relational data. Symbolic ILP models support rule learning in a data-efficient manner. However, symbolic ILP models are not robust to learn from noisy data. Neuro-symbolic ILP models utilize neural networks to learn logic programs in a differentiable manner which improves the robustness of ILP models. However, most neuro-symbolic methods need a strong language bias to learn logic programs, which reduces the usability and flexibility of ILP models and limits the logic program formats. In addition, most neuro-symbolic ILP methods cannot learn logic programs effectively from both small-size datasets and large-size datasets such as knowledge graphs. In the paper, we introduce a novel differentiable ILP model called differentiable first-order rule learner (DFORL), which is scalable to learn rules from both smaller and larger datasets. Besides, DFORL only needs the number of variables in the learned logic programs as input. Hence, DFORL is easy to use and does not need a strong language bias. We demonstrate that DFORL can perform well on several standard ILP datasets, knowledge graphs, and probabilistic relation facts and outperform several well-known differentiable ILP models. Experimental results indicate that DFORL is a precise, robust, scalable, and computationally cheap differentiable ILP model.

Details DOI

IJCAI Conference 2024 Conference Paper

A differentiable first-order rule learner for inductive logic programming (Abstract Reprint)

Kun Gao
Katsumi Inoue
Yongzhi Cao
Hanpin Wang

Learning first-order logic programs from relational facts yields intuitive insights into the data. Inductive logic programming (ILP) models are effective in learning first-order logic programs from observed relational data. Symbolic ILP models support rule learning in a data-ecient manner. However, symbolic ILP models are not robust to learn from noisy data. Neuro-symbolic ILP models utilize neural networks to learn logic programs in a differentiable manner which improves the robustness of ILP models. However, most neuro-symbolic methods need a strong language bias to learn logic programs, which reduces the usability and flexibility of ILP models and limits the logic program formats. In addition, most neuro-symbolic ILP methods cannot learn logic programs effectively from both small-size datasets and large-size datasets such as knowledge graphs. In the paper, we introduce a novel differentiable ILP model called differentiable first-order rule learner (DFORL), which is scalable to learn rules from both smaller and larger datasets. Besides, DFORL only needs the number of variables in the learned logic programs as input. Hence, DFORL is easy to use and does not need a strong language bias. We demonstrate that DFORL can perform well on several standard ILP datasets, knowledge graphs, and probabilistic relation facts and outperform several well-known differentiable ILP models. Experimental results indicate that DFORL is a precise, robust, scalable, and computationally cheap differentiable ILP model.

PDF Details DOI

IJCAI Conference 2022 Conference Paper

Learning First-Order Rules with Differentiable Logic Program Semantics

Kun Gao
Katsumi Inoue
Yongzhi Cao
Hanpin Wang

Learning first-order logic programs (LPs) from relational facts which yields intuitive insights into the data is a challenging topic in neuro-symbolic research. We introduce a novel differentiable inductive logic programming (ILP) model, called differentiable first-order rule learner (DFOL), which finds the correct LPs from relational facts by searching for the interpretable matrix representations of LPs. These interpretable matrices are deemed as trainable tensors in neural networks (NNs). The NNs are devised according to the differentiable semantics of LPs. Specifically, we first adopt a novel propositionalization method that transfers facts to NN-readable vector pairs representing interpretation pairs. We replace the immediate consequence operator with NN constraint functions consisting of algebraic operations and a sigmoid-like activation function. We map the symbolic forward-chained format of LPs into NN constraint functions consisting of operations between subsymbolic vector representations of atoms. By applying gradient descent, the trained well parameters of NNs can be decoded into precise symbolic LPs in forward-chained logic format. We demonstrate that DFOL can perform on several standard ILP datasets, knowledge bases, and probabilistic relation facts and outperform several well-known differentiable ILP models. Experimental results indicate that DFOL is a precise, robust, scalable, and computationally cheap differentiable ILP model.

PDF Details DOI