arXiv subjects

Cheng Zhu

Publications and source records attributed to Cheng Zhu.

27 records · Page 2Linked to original sources

arXiv2023

Controllable Multi-Objective Re-ranking with Policy Hypernetworks

Multi-stage ranking pipelines have become widely used strategies in modern recommender systems, where the final stage aims to return a ranked list of items that balances a number of requirements such as user preference, diversity, novelty etc. Linear scalarization is arguably the most widely used technique to merge multiple requirements into one optimization objective, by summing up the requirements with certain preference weights. Existing final-stage ranking methods often adopt a static model where the preference weights are determined during offline training and kept unchanged during online serving. Whenever a modification of the preference weights is needed, the model has to be re-trained, which is time and resources inefficient. Meanwhile, the most appropriate weights may vary greatly for different groups of targeting users or at different time periods (e.g., during holiday promotions). In this paper, we propose a framework called controllable multi-objective re-ranking (CMR) which incorporates a hypernetwork to generate parameters for a re-ranking model according to different preference weights. In this way, CMR is enabled to adapt the preference weights according to the environment changes in an online manner, without retraining the models. Moreover, we classify practical business-oriented tasks into four main categories and seamlessly incorporate them in a new proposed re-ranking model based on an Actor-Evaluator framework, which serves as a reliable real-world testbed for CMR. Offline experiments based on the dataset collected from Taobao App showed that CMR improved several popular re-ranking models by using them as underlying models. Online A/B tests also demonstrated the effectiveness and trustworthiness of CMR.

cs.IR↗

arXiv2022

Atomic-scale identification of the active sites of nanocatalysts

Alloy nanocatalysts have found broad applications ranging from fuel cells to catalytic converters and hydrogenation reactions. Despite extensive studies, identifying the active sites of nanocatalysts remains a major challenge due to the heterogeneity of the local atomic environment. Here, we advance atomic electron tomography to determine the 3D local atomic structure, surface morphology and chemical composition of PtNi and Mo-doped PtNi nanocatalysts. Using machine learning trained by density functional theory calculations, we identify the catalytic active sites for the oxygen reduction reaction from experimental 3D atomic coordinates, which are corroborated by electrochemical measurements. By quantifying the structure-activity relationship, we discover a local environment descriptor to explain and predict the catalytic active sites at the atomic level. The ability to determine the 3D atomic structure and chemical species coupled with machine learning is expected to expand our fundamental understanding of a wide range of nanocatalysts.

physics.chem-ph↗

arXiv2021

AbdomenCT-1K: Is Abdominal Organ Segmentation A Solved Problem?

With the unprecedented developments in deep learning, automatic segmentation of main abdominal organs seems to be a solved problem as state-of-the-art (SOTA) methods have achieved comparable results with inter-rater variability on many benchmark datasets. However, most of the existing abdominal datasets only contain single-center, single-phase, single-vendor, or single-disease cases, and it is unclear whether the excellent performance can generalize on diverse datasets. This paper presents a large and diverse abdominal CT organ segmentation dataset, termed AbdomenCT-1K, with more than 1000 (1K) CT scans from 12 medical centers, including multi-phase, multi-vendor, and multi-disease cases. Furthermore, we conduct a large-scale study for liver, kidney, spleen, and pancreas segmentation and reveal the unsolved segmentation problems of the SOTA methods, such as the limited generalization ability on distinct medical centers, phases, and unseen diseases. To advance the unsolved problems, we further build four organ segmentation benchmarks for fully supervised, semi-supervised, weakly supervised, and continual learning, which are currently challenging and active research topics. Accordingly, we develop a simple and effective method for each benchmark, which can be used as out-of-the-box methods and strong baselines. We believe the AbdomenCT-1K dataset will promote future in-depth research towards clinical applicable abdominal organ segmentation methods. The datasets, codes, and trained models are publicly available at https://github.com/JunMa11/AbdomenCT-1K.

cs.CV↗

arXiv2020

Learning from Suspected Target: Bootstrapping Performance for Breast Cancer Detection in Mammography

Deep learning object detection algorithm has been widely used in medical image analysis. Currently all the object detection tasks are based on the data annotated with object classes and their bounding boxes. On the other hand, medical images such as mammography usually contain normal regions or objects that are similar to the lesion region, and may be misclassified in the testing stage if they are not taken care of. In this paper, we address such problem by introducing a novel top likelihood loss together with a new sampling procedure to select and train the suspected target regions, as well as proposing a similarity loss to further identify suspected targets from targets. Mean average precision (mAP) according to the predicted targets and specificity, sensitivity, accuracy, AUC values according to classification of patients are adopted for performance comparisons. We firstly test our proposed method on a private dense mammogram dataset. Results show that our proposed method greatly reduce the false positive rate and the specificity is increased by 0.25 on detecting mass type cancer. It is worth mention that dense breast typically has a higher risk for developing breast cancers and also are harder for cancer detection in diagnosis, and our method outperforms a reported result from performance of radiologists. Our method is also validated on the public Digital Database for Screening Mammography (DDSM) dataset, brings significant improvement on mass type cancer detection and outperforms the most state-of-the-art work.

cs.CV↗

arXiv2019

Calibration for computer experiments with binary responses and application to cell adhesion study

Calibration refers to the estimation of unknown parameters which are present in computer experiments but not available in physical experiments. An accurate estimation of these parameters is important because it provides a scientific understanding of the underlying system which is not available in physical experiments. Most of the work in the literature is limited to the analysis of continuous responses. Motivated by a study of cell adhesion experiments, we propose a new calibration framework for binary responses. Its application to the T cell adhesion data provides insight into the unknown values of the kinetic parameters which are difficult to determine by physical experiments due to the limitation of the existing experimental techniques.

stat.ME↗

arXiv2018

A generalized Gaussian process model for computer experiments with binary time series

Non-Gaussian observations such as binary responses are common in some computer experiments. Motivated by the analysis of a class of cell adhesion experiments, we introduce a generalized Gaussian process model for binary responses, which shares some common features with standard GP models. In addition, the proposed model incorporates a flexible mean function that can capture different types of time series structures. Asymptotic properties of the estimators are derived, and an optimal predictor as well as its predictive distribution are constructed. Their performance is examined via two simulation studies. The methodology is applied to study computer simulations for cell adhesion experiments. The fitted model reveals important biological information in repeated cell bindings, which is not directly observable in lab experiments.

stat.ME↗

arXiv2018

Performance Enhancement of Soil-Structure Systems Using a Controlled Rocking

Application of performance-based design (PBD) in earthquake geotechnical design codes has been gaining attention for the past decade. Strong vibrations (e.g., earthquake loading) may generate an uplift or a partial separation of shallow foundations from the underneath soil. To minimize the damage and keep the superstructure safe from the vibrations, a controlled rocking can be exploited. In this study, a finite element method is used to investigate the effect of nonlinear behavior of foundation geomaterial on the rocking behavior of soil-structure systems. To get a better insight about the rocking behavior of the soil-structure system, nonlinear elastic-perfectly plastic behavior is considered for the simulated geomaterial of the foundation. Next, a parametric study, using a wide range of geomaterials with different stiffness values, is conducted in accordance with high-rise structures with different configurations. The results indicate that a significant proportion of the input energy is dissipated by deploying rotational behavior of shallow foundations especially for the structures on softer geomaterials.

physics.geo-ph↗

arXiv2006

Theoretical Study of Sodium and Potassium Resonance Lines Pressure Broadened by Helium Atoms

We perform fully quantum mechanical calculations in the binary approximation of the emission and absorption profiles of the sodium $3s$-$3p$ and potassium $4s$-$4p$ resonance lines under the influence of a helium perturbing gas. We use carefully constructed potential energy surfaces and transition dipole moments to compute the emission and absorption coefficients at temperatures from 158 to 3000 K. Contributions from quasi-bound states are included. The resulting red and blue wing profiles agree well with previous theoretical calculations and with experimental measurements.

physics.atom-ph↗

arXiv2005

Theoretical Study of Pressure Broadening of Lithium Resonance Lines by Helium Atoms

Quantum mechanical calculations are performed of the emission and absorption profiles of the lithium 2s-2p resonance line under the influence of a helium perturbing gas. We use carefully constructed potential energy surfaces and transition dipole moments to compute the emission and absorption coefficients at temperatures from 200 to 3000 K at wavelengths between 500 nm and 1000 nm. Contributions from quasi-bound states are included. The resulting red and blue wing profiles are compared with previous theoretical calculations and with an experiment, carried out at a temperature of 670 K.

physics.atom-ph↗