Author name cluster

Liming Yang

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.

6 papers

1 author row

AAAI Conference 2026 Conference Paper

BAMAS: Structuring Budget-Aware Multi-Agent Systems

Liming Yang
Junyu Luo
Xuanzhe Liu
Yiling Lou
Zhenpeng Chen

Large language model (LLM)-based multi-agent systems have emerged as a powerful paradigm for enabling autonomous agents to solve complex tasks. As these systems scale in complexity, cost becomes an important consideration for practical deployment. However, existing work rarely addresses how to structure multi-agent systems under explicit budget constraints. In this paper, we propose BAMAS, a novel approach for building multi-agent systems with budget awareness. BAMAS first selects an optimal set of LLMs by formulating and solving an Integer Linear Programming problem that balances performance and cost. It then determines how these LLMs should collaborate by leveraging a reinforcement learning-based method to select the interaction topology. Finally, the system is instantiated and executed based on the selected agents and their collaboration topology. We evaluate BAMAS on three representative tasks and compare it with state-of-the-art agent construction methods. Results show that BAMAS achieves comparable performance while reducing cost by up to 86%.

PDF Details DOI

EAAI Journal 2025 Journal Article

Metric learning-based generalized Gaussian kernel for nonlinear classification

Yidan Wang
Liming Yang

Metric learning and kernel learning have attracted a lot of interest for nonlinear classifications in the artificial intelligence field. Most existing nonlinear classification methods are with preset kernel functions, while classification performances depend on kernel functions. To address these shortcomings, we propose a classification framework based on two-stage metric learning. Specifically, in the first stage, a robust metric is trained by applying a robust loss and the maximal margin principle. Then dual technology accompanied by an alternate iteration algorithm is adopted to optimize the problem, and the strong duality property is demonstrated theoretically. The resulting algorithm is convergent. In the second stage, a generalized Gaussian kernel is trained based on the learned metric. This is an adaptive kernel and its main properties are analyzed theoretically. Then with the learned kernel, a nonlinear classification framework is built to train the optimal separation hyperplane, as a classification decision rule. This can avoid effectively the limitation of k-nearest neighbor classifiers. In addition, we analyze the robustness and generalization error bound of the proposed method. Numerical experiments are conducted on two different databases. In the application database, experiments on different spectral regions show that the proposed framework outperforms significantly the traditional methods. Compared with traditional nonlinear methods, the proposed framework increases classification accuracy by an average of 5. 36% in the application database. The statistical tests show the ranking of the proposed framework is the best of all methods. These experiments show the feasibility and effectiveness of the proposed method.

Details DOI

AAAI Conference 2024 Conference Paper

History Matters: Temporal Knowledge Editing in Large Language Model

Xunjian Yin
Jin Jiang
Liming Yang
Xiaojun Wan

The imperative task of revising or updating the knowledge stored within large language models arises from two distinct sources: intrinsic errors inherent in the model which should be corrected and outdated knowledge due to external shifts in the real world which should be updated. Prevailing efforts in model editing conflate these two distinct categories of edits arising from distinct reasons and directly modify the original knowledge in models into new knowledge. However, we argue that preserving the model's original knowledge remains pertinent. Specifically, if a model's knowledge becomes outdated due to evolving worldly dynamics, it should retain recollection of the historical knowledge while integrating the newfound knowledge. In this work, we introduce the task of Temporal Knowledge Editing (TKE) and establish a benchmark AToKe (Assessment of TempOral Knowledge Editing) to evaluate current model editing methods. We find that while existing model editing methods are effective at making models remember new knowledge, the edited model catastrophically forgets historical knowledge. To address this gap, we propose a simple and general framework termed Multi-Editing with Time Objective (METO) for enhancing existing editing models, which edits both historical and new knowledge concurrently and optimizes the model's prediction for the time of each fact. Our assessments demonstrate that while AToKe is still difficult, METO maintains the effectiveness of learning new knowledge and meanwhile substantially improves the performance of edited models on utilizing historical knowledge.

PDF Details DOI

EAAI Journal 2019 Journal Article

A minimax probability extreme machine framework and its application in pattern recognition

Liming Yang
Boyan Yang
Shibo Jing
Qun Sun

In this work we propose a minimax probability extreme learning machine framework (MPME), which combines the benefits of minimax probability machine (MPM) with extreme learning machine (ELM). For binary classification problems, we illustrate that the proposed MPME can be interpreted geometrically by minimizing the maximum of Mahalanobis distances to the two classes. Then two variants of the MPME are presented based on the l 2 -norm loss and l 1 -norm loss functions (called LSEMPME and LADMPME) respectively. Without making specific assumption on the data distribution, the proposed methods can provide explicit upper-bounds for the generalization error, moreover the LSEMPME and LADMPME minimize empirical risk simultaneously. The decision hyperplanes of the proposed methods pass through the origin in ELM feature space with few decision variables. By using the multivariate Chebyshev–Cantelli inequality, all the proposed problems can be reformulated as second-order cone programming (SOCP) with global solutions. Furthermore, numerical experiments have been carried out on two databases that are drawn from UCI benchmark database and a practical application database. First, the proposed methods are evaluated for a practical application consisting on the analysis of licorice seeds using near-infrared spectral (NIR) data. Experiments in six different spectral regions illustrate that the proposed methods can improve generalization in most cases. Then the proposed methods are evaluated on benchmark datasets. In comparison with traditional methods including MPM, ELM and support vector machine (SVM), experiments show that the proposed methods achieve comparable results in generalization. With few decision variables, the proposed methods are easy to implement for nonlinear classification and to estimate a lower-bound on the prediction accuracy.

Details DOI

EAAI Journal 2016 Journal Article

A sparse extreme learning machine framework by continuous optimization algorithms and its application in pattern recognition

Liming Yang
Siyun Zhang

Extreme learning machine (ELM) has demonstrated great potential in machine learning owing to its simplicity, rapidity and good generalization performance. In this investigation, based on least-squares estimate (LSE) and least absolute deviation (LAD), we propose four sparse ELM formulations with zero-norm regularization to automatically choose the optimal hidden nodes. Furthermore, we develop two continuous optimization methods to solve the proposed problems respectively. The first is DC (difference of convex functions) approximation approach that approximates the zero-norm by a DC function, and the resulting optimizations are posed as DC programs. The second is an exact penalty technique for zero-norm, and the resulting problems are reformulated as DC programs, and the corresponding DCAs converge finitely. Moreover, the proposed framework is applied directly to recognize the hardness of licorice seeds using near-infrared spectral data. Experiments in different spectral regions illustrate that the proposed approaches can reduce the number of hidden nodes (or output features), while either improve or show no significant difference in generalization compared with the traditional ELM methods and support vector machine (SVM). Experiments on several benchmark data sets demonstrate that the proposed framework is competitive with the traditional approaches in generalization, but selects fewer output features.

Details DOI

EAAI Journal 2014 Journal Article

A convex relaxation framework for a class of semi-supervised learning methods and its application in pattern recognition

Liming Yang
Laisheng Wang
Yongping Gao
Qun Sun
Tengyang Zhao

Semi-supervised learning has been an attractive research tool for using unlabeled data in pattern recognition. Applying a novel semi-definite programming (SDP) relaxation strategy to a class of continuous semi-supervised support vector machines (S3VMs), a new convex relaxation framework for the S3VMs is proposed based on SDP. Compared with other SDP relaxations for S3VMs, the proposed methods only require solving the primal problems and can implement L 1-norm regularization. Furthermore, the proposed technique is applied directly to recognize the purity of hybrid maize seeds using near-infrared spectral data, from which we find that the proposed method achieves equivalent performance to the exact solution algorithm for solving the S3VM in different spectral regions. Experiments on several benchmark data sets demonstrate that the proposed convex technique is competitive with other SDP relaxation methods for solving semi-supervised SVMs in generalization.

Details DOI