Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileAcademic lineage
Possible advisorsa guess from early papers, not confirmed
- Kian Hsiang LowSuggested from co-authorship
Is this you? Claim this profile to confirm or dismiss it.
Works72 from public data
- 141
- A Unifying Framework of Anytime Sparse Gaussian Process Regression Models with Stochastic Variational Inference for Big Data95
This paper presents a novel unifying framework of anytime sparse Gaussian process regression (SGPR) models that can produce good predictive performance fast and improve their predictive performance over time and empirically evaluates the trade-off between the predictive performance vs. time efficiency of the anytime SGPR models on two real-world million-sized datasets.
- Verifiably safe exploration for end-to-end reinforcement learning67
A first approach toward enforcing formal safety constraints on end-to-end policies with visual inputs is contributed, drawing on recent advances in object detection and automated reasoning for hybrid dynamical systems.
- A distributed variational inference framework for unifying parallel sparse Gaussian process regression models60
A novel distributed variational inference framework that unifies many parallel sparse Gaussian process regression (SGPR) models for scalable hyperparameter learning with big data and empirically evaluates the predictive performance and scalability of the distributedvariational SGPR models unified by this framework on two real-world datasets.
- Decentralized High-Dimensional Bayesian Optimization With Factor Graphs59
A novel decentralized high-dimensional Bayesian optimization algorithm that can exploit the interdependent effects of various input components on the output of the unknown objective function f for boosting the BO performance and still preserve scalability in the number of input dimensions without requiring prior knowledge or the existence of a low dimension of the input space is presented.
- Active Learning Is Planning: Nonmyopic ε-Bayes-Optimal Active Learning of Gaussian Processes50
This paper presents a novel nonmyopic e-Bayes-optimal active learning (e-BAL) approach that jointly optimizes the exploration-exploitation trade-off and proposes an anytime algorithm based on eBAL with performance guarantee and empirically demonstrates that it outperforms the state-of-the-art algorithms.
- Collective Online Learning of Gaussian Processes in Massive Multi-Agent Systems49
A rigorous theoretical analysis of the approximation loss arising from the proposed representation to achieve efficient online updates and model fusion and shows that COOL-GP is highly effective in model fusion, resilient to information disparity between agents, robust to transmission loss, and can scale to thousands of agents.
- Near-Optimal Active Learning of Multi-Output Gaussian Processes42
This paper exploits a structure common to sparse MOGP models for deriving a novel active learning criterion and exploits a relaxed form of submodularity property of the new criterion for devising a polynomial-time approximation algorithm that guarantees a constant-factor approximation of that achieved by the optimal set of selected observations.
- A Generalized Stochastic Variational Bayesian Hyperparameter Learning Framework for Sparse Spectrum Gaussian Process Regression40
The generalized framework of stochastic variational Bayesian sparse spectrum GP (sVBSSGP) models addresses their shortcomings by adopting a Bayesian treatment of the spectral frequencies to avoid overfitting, and exploiting local data for boosting the predictive performance.
- Nonmyopic \epsilon-Bayes-Optimal Active Learning of Gaussian Processes38
To perform active learning in real time, this paper proposes an anytime algorithm based on e-BAL with performance guarantee and empirically demonstrate that, with limited budget, it outperforms the state-of-the-art algorithms.
- DDL: Deep Dictionary Learning for Predictive Phenotyping35
A deep dictionary learning framework (DDL) for phenotyping is proposed, which utilizes unlabeled data as a complementary source of information to generate a better, more succinct data representation that helps improve DDL’s phenotyping performance over existing methods that only uses labeled data.
- Collective Model Fusion for Multiple Black-Box Experts35
This paper presents the first collective model fusion framework for multiple experts with heterogeneous black-box architectures, addressing the key issues of how black- box experts interact to understand the predictive behaviors of one another and how the shared understandings can be combined to generate high-quality consensus prediction.
- 31
- 30
- Offline Model-Based Optimization via Policy-Guided Gradient Search25
This work introduces a new learning-to-search perspective for offline optimization by reformulating it as an offline reinforcement learning problem and explicitly learns the best policy for a given surrogate model created from the offline data.
- Learning Task-Agnostic Embedding of Multiple Black-Box Experts for Multi-Task Model Fusion25
A new fusion paradigm is developed that represents each expert as a distribution over a spectrum of predictive prototypes, which are isolated from task-specific information encoded within the prototype distribution and can then be reintegrated to generate a new model that solves a new task encoded with a different prototype distribution.
- Recent Advances in Scaling Up Gaussian Process Predictive Models for Large Spatiotemporal Data22
An overview of the recent progress in scaling up GP models for large spatiotemporally correlated data through parallelization on clusters of machines, online learning, and nonmyopic active sensing/learning is presented.
- Probabilistic Federated Prompt-Tuning with Non-IID and Imbalanced Data19
This work transforms federated learning into a distributed set modeling task, aggregating diverse sets of prompts to globally fine-tune the pre-trained model, and introduces a new probabilistic prompt aggregation method that substantially outperforms existing federated model aggregation techniques.
- Scalable Decision-Theoretic Coordination and Control for Real-time Active Multi-Camera Surveillance18
An overview of the novel decision-theoretic multi-agent approach for controlling and coordinating multiple active cameras in surveillance, which focuses on two novel surveillance tasks: maximize the number of targets observed in active cameras with guaranteed image resolution and to improve the fairness in observation of multiple targets.
- AID: Active Distillation Machine to Leverage Pre-Trained Black-Box Models in Private Data Settings13
This paper presents an active distillation method for a local institution to find the best queries within its given budget to distill an on-server black-box model’s predictive knowledge into a local surrogate with transparent parameterization, which addresses several challenges of deploying machine learning in many industrial settings with strong proprietary constraints.
- 13
- 13
- Boosting Offline Optimizers with Surrogate Sensitivity12
An optimizable sensitivity measurement is developed for the surrogate model, which inspires a sensitivity-informed regularizer that is applicable to a wide range of offline optimizers and is demonstrated in the extensive experiment benchmark.
- Incentives in Private Collaborative Machine Learning12
This work introduces differential privacy (DP) as an incentive, which enforces a privacy-valuation trade-off, and rewards each party with different posterior samples of the model parameters.
- Near-Optimal Adversarial Policy Switching for Decentralized Asynchronous Multi-Agent Systems11
The approach first optimizes a set of stratagems that represent best responses to detect changes in adversary strategies and then integrated into a unified policy that can detect and respond when the adversaries change their strategies.
- 11
- Federated Learning of Models Pre-Trained on Different Features with Consensus Graphs10
A feature fusion approach is proposed that extracts local representations from local models and incorporates them into a global representation that improves the prediction performance and is demonstrated in real-world applications on time series data such as power grids and traffic networks.
- 10
- Pre-trained Recommender Systems: A Causal Debiasing Perspective9
This paper proposes to develop a generic recommender that captures universal interaction patterns by training on generic user-item interaction data extracted from different domains, which can then be fast adapted to improve few-shot learning performance in unseen new domains (with limited data).
- Personalized Federated Domain Adaptation for Item-to-Item Recommendation9
A personalized graph adaptation model is proposed that bridges the gap between recent literature on federated GNNs and (non-graph) personalized federated learning, which either does not optimize for the adaptability of the federated model or is restricted to local models with homogeneous parameterization.
- 9
- Incorporating Surrogate Gradient Norm to Improve Offline Optimization Techniques8
A model-agnostic approach to offline optimization that incorporates a notion of model sharpness into the training loss of the surrogate as a regularizer, supported by a new theoretical analysis demonstrating that reducing surrogate sharpness on the offline dataset provably reduces its generalized sharpness on unseen data.
- 8
- Federated Prompt-Tuning with Heterogeneous and Incomplete Multimodal Client Data7
A generalized federated prompttuning framework for practical scenarios where local datasets are multi-modal and exhibit different distributional patterns of missing features at the input level, which allows prompt instructions to complement one another and be combined effectively.
- Few-Shot Learning via Repurposing Ensemble of Black-Box Models6
This paper investigates the problem of exploiting existing solution models of previous tasks to address a related target task with limited training data from a new perspective of black-box re-purposing, which augments the target inputs and leverages their corresponding outputs generated by existing black-box APIs into a feature ensemble.
- 6
- Adaptive Multi-Source Causal Inference from Observational Data5
This work proposes a new approach to estimate causal effects from observational data that leverages multiple data sources which share similar causal mechanisms with the scarce target observations to help infer causal effects in the target domain.
- Learning Personalized Item-to-Item Recommendation Metric via Implicit Feedback5
This paper develops and investigates a personalizable deep metric model that captures both the internal contents of items and how they were interacted with by users, and develops a new metric representation embedded as kernel parameters of a probabilistic model.
- Collective Online Learning via Decentralized Gaussian Processes in Massive Multi-Agent Systems.5
This paper introduces a novel Collective Online Learning Gaussian Process framework for massive distributed systems that allows each agent to build its local model, which can be exchanged and combined efficiently with others via peer-to-peer communication to converge on a global model of higher quality.
- Stochastic Variational Inference for Fully Bayesian Sparse Gaussian Process Regression Models.5
It is shown that the stochastic gradient is an unbiased estimator of the exact gradient and can be computed in constant time per iteration, hence achieving scalability to big data.
- ROOT: Rethinking Offline Optimization as Distributional Translation via Probabilistic Bridge4
The black-box optimization task which aims to find the maxima of a black-box function using a static set of its observed input-output pairs is studied, with a new perspective that casts offline optimization as a distributional translation task.
- Robust Multivariate Time-Series Forecasting: Adversarial Attacks and Defense Mechanisms4
This work studies and design adversarial attack on multivariate probabilistic forecasting models, taking into consideration attack budget constraints and the correlation architecture between multiple time series, and develops two defense strategies.
- 3
- CHEER: Rich Model Helps Poor Model via Knowledge Infusion3
A knowledge infusion framework named <monospace>CHEER</ monospace> is developed that can succinctly summarize such rich model into transferable representations, which can be incorporated into the poor model to improve its performance.
- Effective knowledge representation and utilization for sustainable collaborative learning across heterogeneous systems2
A new research agenda is proposed on developing effective and sustainable collaborative learning frameworks across heterogeneous systems, featuring three novel computational capabilities on knowledge organization: model expression, comprehension, and valuation.
- 2
- 2
- 2
- A Multi-agent Mechanism in Machine Learning Approach to Anti-virus System2
A multi-agent mechanism to protect target systems from computer virus infections is introduced and a form of object knowledge is defined to determine computer viruses and diagnosed objects and an association model of knowledge base and database is set up.
- Leveraging Soft Prompts for Privacy Attacks in Federated Prompt Tuning1
This work proposes PromptMIA, a membership inference attack tailored to federated prompt-tuning, in which a malicious server introduces adversarially crafted prompts and exploits their updates during collaborative training to determine whether a target data point belongs to a client's private dataset.
- Rethinking Cross-Modal Fine-Tuning: Optimizing the Interaction Between Feature Alignment and Target Fitting1
A principled framework is developed that establishes a provable generalization bound on the target error, which explains the interaction between feature alignment and target fitting through a novel concept of feature-label distortion.
- 1
- 1
- 1
- 1
- Intention-Aware Planning under Uncertainty for Interacting with Self-Interested, Boundedly Rational Agents (Extended Abstract)1
This work investigates how intention prediction can be efficiently exploited and made practical in planning, thereby leading to efficient intention-aware planning frameworks capable of predicting the intentions of other agents and acting optimally with respect to their predicted intentions.
- Intention-aware planning under uncertainty for interacting with self-interested, boundedly rational agents1
This work investigates how intention prediction can be efficiently exploited and made practical in planning, thereby leading to efficient intention-aware planning frameworks capable of predicting the intentions of other agents and acting optimally with respect to their predicted intentions.
- Diffusion-Inspired Reconfiguration of Transformers for Uncertainty Calibration–
A diffusion-inspired reconfiguration of transformers in which each feature transformation block is modeled as a probabilistic mapping that mimics the structure of a diffusion process, transporting data mass from the input distribution to the pre-trained feature distribution.
- D-FROST: Decentralized Federated pRompt-tuning via Optimal tranSporT for Non-IID and Imbalanced Data–
- –
- –
- Cross-Modal Knowledge Distillation without Paired Data: Theoretical Foundation and Algorithm–
A principled framework is proposed that enables effective cross-modal knowledge distillation by aligning distributions rather than individual samples, and is highly effective in both unpaired and paired data settings, improving significantly over prior work.
- Black-Box Optimization From Small Offline Datasets via Meta Learning with Synthetic Tasks–
OptBias is proposed, a meta-learning framework that directly tackles data scarcity that learns a reusable optimization bias by training on synthetic tasks generated from a Gaussian process, and then fine-tunes the surrogate model on the small data for the target task.
- Causal Graph Learning via Distributional Invariance of Cause-Effect Relationship–
An algorithm is developed that efficiently uncovers causal relationships with quadratic complexity in the number of observational variables, reducing the processing time by up to 25x compared to state-of-the-art methods.
- FlowTransOP: Distributional Translation of Omics Signatures via Constrained Deep Flow Matching–
flow matching is used to align full domain distributions in a pre-aligned latent space, with a structural regularization term that keeps similar conditions proximate after transformation, enabling reliable therapeutic inference.
- Model Merging as Probabilistic Inference in Fine-Tuning Parameter Space–
A heavy-tailed PoE design based on Cauchy experts is presented, which better captures the observed residual behavior while admitting a provably convergent inference procedure.
- –
- –
- FedCert: Federated Accuracy Certification–
The proposed FedCert method is designed to approximate the certified accuracy of a global model based on the certified accuracy and class distribution of each client, and introduces the client grouping algorithm to ensure reliable certified accuracy during the aggregation step of the approximation algorithm.
- Time-to-Pattern: Information-Theoretic Unsupervised Learning for Scalable Time Series Summarization–
This work introduces an approach to time series summarization, called Time-to-Pattern (T2P), which aims to find a set of diverse patterns that together encode the most salient information, following the notion of minimum description length.
- –
- –
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-12. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.