Martin Ester
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Martin Ester has a total of 160 co-authors. Here are the top co-authors by number of shared publications:
- Artem Cherkasov
The Vancouver Prostate Centre, Department of Urologic Sciences, University of British Columbia, Vancouver, Canada
6 shared publications - Colin Collins
Vancouver Prostate Centre, Vancouver, BC V6H 3Z6, Canada
6 shared publications - Olga Zolotareva
Institute for Computational Systems Biomedicine, University of Hamburg, Hamburg, Germany
4 shared publications Hossein Sharifi-Noghabi
School of Computing Science, Simon Fraser University, Burnaby, British Columbia, Canada
3 shared publications- Jiajun Bu
Zhejiang Key Lab of Accessible Perception & Intelligent Systems, Zhejiang University, Hangzhou, China
3 shared publications - Xin Wang
Tsinghua University, Beijing, China
3 shared publications
Latest Publications
- Open Access
Why Pool When You Can Flow? Active Learning with GFlowNets
The scalability of pool-based active learning is limited by the computational cost of evaluating large unlabeled datasets, a challenge that is particularly acute in virtual screening for drug discovery. While active learning strategies such as Bayesian Active Learning by Disagreement (BALD) prioritize informative samples, it remains computationally intensive when scaled to libraries containing billions samples. In this work, we introduce BALD-GFlowNet, a generative active learning framework that circumvents this issue. Our method leverages Generative Flow Networks (GFlowNets) to directly sample objects in proportion to the BALD reward. By replacing traditional pool-based acquisition with generative sampling, BALD-GFlowNet achieves scalability that is independent of the size of the unlabeled pool. In our virtual screening experiment, we show that BALD-GFlowNet achieves a performance comparable to that of standard BALD baseline while generating more structurally diverse molecules, offering a promising direction for efficient and scalable molecular discovery.
Open AccessWhy Pool When You Can Flow? Active Learning with GFlowNets
The scalability of pool-based active learning is limited by the computational cost of evaluating large unlabeled datasets, a challenge that is particularly acute in virtual screening for drug discovery. While active learning strategies such as Bayesian Active Learning by Disagreement (BALD) prioritize informative samples, it remains computationally intensive when scaled to libraries containing billions samples. In this work, we introduce BALD-GFlowNet, a generative active learning framework that circumvents this issue. Our method leverages Generative Flow Networks (GFlowNets) to directly sample objects in proportion to the BALD reward. By replacing traditional pool-based acquisition with generative sampling, BALD-GFlowNet achieves scalability that is independent of the size of the unlabeled pool. In our virtual screening experiment, we show that BALD-GFlowNet achieves a performance comparable to that of standard BALD baseline while generating more structurally diverse molecules, offering a promising direction for efficient and scalable molecular discovery.
- 15 June 2025
- Published by Springer Nature
- p. 306-318
Informed Augmentation Selection Improves Tabular Contrastive Learning
- 15 June 2025
- Published by Springer Nature
- p. 306-318
Informed Augmentation Selection Improves Tabular Contrastive Learning
- Open Access
Improving OOD Generalization of Pre-trained Encoders via Aligned Embedding-Space Ensembles
The quality of self-supervised pre-trained embeddings on out-of-distribution (OOD) data is poor without fine-tuning. A straightforward and simple approach to improving the generalization of pre-trained representation to OOD data is the use of deep ensembles. However, obtaining an effective ensemble in the embedding space with only unlabeled data remains an unsolved problem. We first perform a theoretical analysis that reveals the relationship between individual hyperspherical embedding spaces in an ensemble. We then design a principled method to align these embedding spaces in an unsupervised manner. Experimental results on the MNIST dataset show that our embedding-space ensemble method improves pre-trained embedding quality on in-distribution and OOD data compared to single encoders.
Open AccessImproving OOD Generalization of Pre-trained Encoders via Aligned Embedding-Space Ensembles
The quality of self-supervised pre-trained embeddings on out-of-distribution (OOD) data is poor without fine-tuning. A straightforward and simple approach to improving the generalization of pre-trained representation to OOD data is the use of deep ensembles. However, obtaining an effective ensemble in the embedding space with only unlabeled data remains an unsolved problem. We first perform a theoretical analysis that reveals the relationship between individual hyperspherical embedding spaces in an ensemble. We then design a principled method to align these embedding spaces in an unsupervised manner. Experimental results on the MNIST dataset show that our embedding-space ensemble method improves pre-trained embedding quality on in-distribution and OOD data compared to single encoders.
- Top Cited PapersOpen Access
- 11 November 2024
- Published by Association for Computing Machinery (ACM) in ACM Computing Surveys
- Vol. 57 (3) , 1-38
A Comprehensive Survey on Deep Clustering: Taxonomy, Challenges, and Future Directions
Clustering is a fundamental machine learning task, which aim at assigning instances into groups so that similar samples belong to the same cluster while dissimilar samples belong to different clusters. Shallow clustering methods usually assume that data are collected and expressed as feature vectors within which clustering is performed. However, clustering high-dimensional data, such as images, texts, videos, and graphs, poses significant challenges for clustering tasks, such as indiscriminate representation and intricate relationships among instances. Over the past decades, deep learning has achieved remarkable success in effective representation learning and modeling complex relationships. Motivated by these advancements, Deep Clustering seeks to improve clustering outcomes through deep learning techniques, garnering considerable interest from both academia and industry. Despite many contributions to this vibrant area of research, the lack of systematic analysis and a comprehensive taxonomy has hindered progress in this field. In this survey, we first explore how deep learning can be integrated into deep clustering and identify two fundamental components: the representation learning module and the clustering module. Then, we summarize and analyze the representative design of these two modules. Furthermore, we introduce a novel taxonomy of deep clustering based on how these two modules interact, specifically through multistage, generative, iterative, and simultaneous approaches. In addition, we present well-known benchmark datasets, evaluation metrics, and open-source tools to clearly demonstrate different experimental approaches. Finally, we examine the practical applications of deep clustering and propose challenging areas for future research.
- 11 November 2024
- Published by Association for Computing Machinery (ACM) in ACM Computing Surveys
- Vol. 57 (3) , 1-38
Top Cited PapersOpen AccessA Comprehensive Survey on Deep Clustering: Taxonomy, Challenges, and Future Directions
Clustering is a fundamental machine learning task, which aim at assigning instances into groups so that similar samples belong to the same cluster while dissimilar samples belong to different clusters. Shallow clustering methods usually assume that data are collected and expressed as feature vectors within which clustering is performed. However, clustering high-dimensional data, such as images, texts, videos, and graphs, poses significant challenges for clustering tasks, such as indiscriminate representation and intricate relationships among instances. Over the past decades, deep learning has achieved remarkable success in effective representation learning and modeling complex relationships. Motivated by these advancements, Deep Clustering seeks to improve clustering outcomes through deep learning techniques, garnering considerable interest from both academia and industry. Despite many contributions to this vibrant area of research, the lack of systematic analysis and a comprehensive taxonomy has hindered progress in this field. In this survey, we first explore how deep learning can be integrated into deep clustering and identify two fundamental components: the representation learning module and the clustering module. Then, we summarize and analyze the representative design of these two modules. Furthermore, we introduce a novel taxonomy of deep clustering based on how these two modules interact, specifically through multistage, generative, iterative, and simultaneous approaches. In addition, we present well-known benchmark datasets, evaluation metrics, and open-source tools to clearly demonstrate different experimental approaches. Finally, we examine the practical applications of deep clustering and propose challenging areas for future research.
- Open Access
- 1 November 2024
- Published by Elsevier in Computers in Biology and Medicine
- Vol. 183, 109324
Machine learning model identifies genetic predictors of cisplatin-induced ototoxicity in CERS6 and TLR4
- 1 November 2024
- Published by Elsevier in Computers in Biology and Medicine
- Vol. 183, 109324
Open AccessMachine learning model identifies genetic predictors of cisplatin-induced ototoxicity in CERS6 and TLR4
- Open Access
Causal Order Discovery based on Monotonic SCMs
In this paper, we consider the problem of causal order discovery within the framework of monotonic Structural Causal Models (SCMs), which have gained attention for their potential to enable causal inference and causal discovery from observational data. While existing approaches either assume prior knowledge about the causal order or use complex optimization techniques to impose sparsity in the Jacobian of Triangular Monotonic Increasing maps, our work introduces a novel sequential procedure that directly identifies the causal order by iteratively detecting the root variable. This method eliminates the need for sparsity assumptions and the associated optimization challenges, enabling the identification of a unique SCM without the need for multiple independence tests to break the Markov equivalence class. We demonstrate the effectiveness of our approach in sequentially finding the root variable, comparing it to methods that maximize Jacobian sparsity.
Open AccessCausal Order Discovery based on Monotonic SCMs
In this paper, we consider the problem of causal order discovery within the framework of monotonic Structural Causal Models (SCMs), which have gained attention for their potential to enable causal inference and causal discovery from observational data. While existing approaches either assume prior knowledge about the causal order or use complex optimization techniques to impose sparsity in the Jacobian of Triangular Monotonic Increasing maps, our work introduces a novel sequential procedure that directly identifies the causal order by iteratively detecting the root variable. This method eliminates the need for sparsity assumptions and the associated optimization challenges, enabling the identification of a unique SCM without the need for multiple independence tests to break the Markov equivalence class. We demonstrate the effectiveness of our approach in sequentially finding the root variable, comparing it to methods that maximize Jacobian sparsity.
- Open Access
IceFormer: Accelerated Inference with Long-Sequence Transformers on CPUs
One limitation of existing Transformer-based models is that they cannot handle very long sequences as input since their self-attention operations exhibit quadratic time and space complexity. This problem becomes especially acute when Transformers are deployed on hardware platforms equipped only with CPUs. To address this issue, we propose a novel method for accelerating self-attention at inference time that works with pretrained Transformer models out-of-the-box without requiring retraining. We experiment using our method to accelerate various long-sequence Transformers, including a leading LLaMA 2-based LLM, on various benchmarks and demonstrate a greater speedup of 2.73x - 7.63x while retaining 98.6% - 99.6% of the accuracy of the original pretrained models. The code is available on our project website at https://yuzhenmao.github.io/IceFormer/.
Open AccessIceFormer: Accelerated Inference with Long-Sequence Transformers on CPUs
One limitation of existing Transformer-based models is that they cannot handle very long sequences as input since their self-attention operations exhibit quadratic time and space complexity. This problem becomes especially acute when Transformers are deployed on hardware platforms equipped only with CPUs. To address this issue, we propose a novel method for accelerating self-attention at inference time that works with pretrained Transformer models out-of-the-box without requiring retraining. We experiment using our method to accelerate various long-sequence Transformers, including a leading LLaMA 2-based LLM, on various benchmarks and demonstrate a greater speedup of 2.73x - 7.63x while retaining 98.6% - 99.6% of the accuracy of the original pretrained models. The code is available on our project website at https://yuzhenmao.github.io/IceFormer/.
- Open Access
- 24 March 2024
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 38 (12) , 13085-13093
Adversarially Balanced Representation for Continuous Treatment Effect Estimation
Individual treatment effect (ITE) estimation requires adjusting for the covariate shift between populations with different treatments, and deep representation learning has shown great promise in learning a balanced representation of covariates. However the existing methods mostly consider the scenario of binary treatments. In this paper, we consider the more practical and challenging scenario in which the treatment is a continuous variable (e.g. dosage of a medication), and we address the two main challenges of this setup. We propose the adversarial counterfactual regression network (ACFR) that adversarially minimizes the representation imbalance in terms of KL divergence, and also maintains the impact of the treatment value on the outcome prediction by leveraging an attention mechanism. Theoretically we demonstrate that ACFR objective function is grounded in an upper bound on counterfactual outcome prediction error. Our experimental evaluation on semi-synthetic datasets demonstrates the empirical superiority of ACFR over a range of state-of-the-art methods.
- 24 March 2024
- Published by Association for the Advancement of Artificial Intelligence (AAAI) in Proceedings of the AAAI Conference on Artificial Intelligence
- Vol. 38 (12) , 13085-13093
Open AccessAdversarially Balanced Representation for Continuous Treatment Effect Estimation
Individual treatment effect (ITE) estimation requires adjusting for the covariate shift between populations with different treatments, and deep representation learning has shown great promise in learning a balanced representation of covariates. However the existing methods mostly consider the scenario of binary treatments. In this paper, we consider the more practical and challenging scenario in which the treatment is a continuous variable (e.g. dosage of a medication), and we address the two main challenges of this setup. We propose the adversarial counterfactual regression network (ACFR) that adversarially minimizes the representation imbalance in terms of KL divergence, and also maintains the impact of the treatment value on the outcome prediction by leveraging an attention mechanism. Theoretically we demonstrate that ACFR objective function is grounded in an upper bound on counterfactual outcome prediction error. Our experimental evaluation on semi-synthetic datasets demonstrates the empirical superiority of ACFR over a range of state-of-the-art methods.
- Open Access
- 1 February 2024
- Published by Oxford University Press (OUP) in Bioinformatics
- Vol. 40 (2)
Phenotype prediction from single-cell RNA-seq data using attention-based neural networks
Motivation: A patient’s disease phenotype can be driven and determined by specific groups of cells whose marker genes are either unknown or can only be detected at late-stage using conventional bulk assays such as RNA-Seq technology. Recent advances in single-cell RNA sequencing (scRNA-seq) enable gene expression profiling in cell-level resolution, and therefore have the potential to identify those cells driving the disease phenotype even while the number of these cells is small. However, most existing methods rely heavily on accurate cell type detection, and the number of available annotated samples is usually too small for training deep learning predictive models. Results: Here, we propose the method ScRAT for phenotype prediction using scRNA-seq data. To train ScRAT with a limited number of samples of different phenotypes, such as coronavirus disease (COVID) and non-COVID, ScRAT first applies a mixup module to increase the number of training samples. A multi-head attention mechanism is employed to learn the most informative cells for each phenotype without relying on a given cell type annotation. Using three public COVID datasets, we show that ScRAT outperforms other phenotype prediction methods. The performance edge of ScRAT over its competitors increases as the number of training samples decreases, indicating the efficacy of our sample mixup. Critical cell types detected based on high-attention cells also support novel findings in the original papers and the recent literature. This suggests that ScRAT overcomes the challenge of missing marker genes and limited sample number with great potential revealing novel molecular mechanisms and/or therapies. Availability and implementation: The code of our proposed method ScRAT is published at https://github.com/yuzhenmao/ScRAT.
- 1 February 2024
- Published by Oxford University Press (OUP) in Bioinformatics
- Vol. 40 (2)
Open AccessPhenotype prediction from single-cell RNA-seq data using attention-based neural networks
Motivation: A patient’s disease phenotype can be driven and determined by specific groups of cells whose marker genes are either unknown or can only be detected at late-stage using conventional bulk assays such as RNA-Seq technology. Recent advances in single-cell RNA sequencing (scRNA-seq) enable gene expression profiling in cell-level resolution, and therefore have the potential to identify those cells driving the disease phenotype even while the number of these cells is small. However, most existing methods rely heavily on accurate cell type detection, and the number of available annotated samples is usually too small for training deep learning predictive models. Results: Here, we propose the method ScRAT for phenotype prediction using scRNA-seq data. To train ScRAT with a limited number of samples of different phenotypes, such as coronavirus disease (COVID) and non-COVID, ScRAT first applies a mixup module to increase the number of training samples. A multi-head attention mechanism is employed to learn the most informative cells for each phenotype without relying on a given cell type annotation. Using three public COVID datasets, we show that ScRAT outperforms other phenotype prediction methods. The performance edge of ScRAT over its competitors increases as the number of training samples decreases, indicating the efficacy of our sample mixup. Critical cell types detected based on high-attention cells also support novel findings in the original papers and the recent literature. This suggests that ScRAT overcomes the challenge of missing marker genes and limited sample number with great potential revealing novel molecular mechanisms and/or therapies. Availability and implementation: The code of our proposed method ScRAT is published at https://github.com/yuzhenmao/ScRAT.
- Open Access
Adversarially Balanced Representation for Continuous Treatment Effect Estimation
Individual treatment effect (ITE) estimation requires adjusting for the covariate shift between populations with different treatments, and deep representation learning has shown great promise in learning a balanced representation of covariates. However the existing methods mostly consider the scenario of binary treatments. In this paper, we consider the more practical and challenging scenario in which the treatment is a continuous variable (e.g. dosage of a medication), and we address the two main challenges of this setup. We propose the adversarial counterfactual regression network (ACFR) that adversarially minimizes the representation imbalance in terms of KL divergence, and also maintains the impact of the treatment value on the outcome prediction by leveraging an attention mechanism. Theoretically we demonstrate that ACFR objective function is grounded in an upper bound on counterfactual outcome prediction error. Our experimental evaluation on semi-synthetic datasets demonstrates the empirical superiority of ACFR over a range of state-of-the-art methods.
Open AccessAdversarially Balanced Representation for Continuous Treatment Effect Estimation
Individual treatment effect (ITE) estimation requires adjusting for the covariate shift between populations with different treatments, and deep representation learning has shown great promise in learning a balanced representation of covariates. However the existing methods mostly consider the scenario of binary treatments. In this paper, we consider the more practical and challenging scenario in which the treatment is a continuous variable (e.g. dosage of a medication), and we address the two main challenges of this setup. We propose the adversarial counterfactual regression network (ACFR) that adversarially minimizes the representation imbalance in terms of KL divergence, and also maintains the impact of the treatment value on the outcome prediction by leveraging an attention mechanism. Theoretically we demonstrate that ACFR objective function is grounded in an upper bound on counterfactual outcome prediction error. Our experimental evaluation on semi-synthetic datasets demonstrates the empirical superiority of ACFR over a range of state-of-the-art methods.