User profiles for Zhaowei Zhu
Zhaowei ZhuUniversity of California, Santa Cruz Verified email at ucsc.edu Cited by 1038 |
Learning with noisy labels revisited: A study using real-world human annotations
Existing research on learning with noisy labels mainly focuses on synthetic label noise.
Synthetic noise, though has clean structures which greatly enabled statistical analyses, often …
Synthetic noise, though has clean structures which greatly enabled statistical analyses, often …
Learning with instance-dependent label noise: A sample sieve approach
Human-annotated labels are often prone to noise, and the presence of such noise will
degrade the performance of the resulting deep neural network (DNN) models. Much of the …
degrade the performance of the resulting deep neural network (DNN) models. Much of the …
A second-order approach to learning with instance-dependent label noise
The presence of label noise often misleads the training of deep neural networks. Departing
from the recent literature which largely assumes the label noise rate is only determined by …
from the recent literature which largely assumes the label noise rate is only determined by …
Effect of platelet‐rich plasma (PRP) concentration on proliferation, neurotrophic function and migration of Schwann cells in vitro
C Zheng, Q Zhu, X Liu, X Huang, C He… - Journal of tissue …, 2016 - Wiley Online Library
Platelet‐rich plasma (PRP) contains various growth factors and appears to have the potential
to promote peripheral nerve regeneration, but evidence is lacking regarding its biological …
to promote peripheral nerve regeneration, but evidence is lacking regarding its biological …
[HTML][HTML] Factors predicting sensory and motor recovery after the repair of upper limb peripheral nerve injuries
B He, Z Zhu, Q Zhu, X Zhou, C Zheng, P Li… - Neural regeneration …, 2014 - journals.lww.com
… Bo He and Zhaowei Zhu contributed equally to this work. … Author contributions:Liu XL, Zhu
JK and He B were responsible for the study design. Zheng CB and Li PL collected and input …
JK and He B were responsible for the study design. Zheng CB and Li PL collected and input …
Detecting corrupted labels without training a model to predict
Label noise in real-world datasets encodes wrong correlation patterns and impairs the
generalization of deep neural networks (DNNs). It is critical to find efficient ways to detect …
generalization of deep neural networks (DNNs). It is critical to find efficient ways to detect …
Clusterability as an alternative to anchor points when learning with noisy labels
The label noise transition matrix, characterizing the probabilities of a training instance being
wrongly annotated, is crucial to designing popular solutions to learning with noisy labels. …
wrongly annotated, is crucial to designing popular solutions to learning with noisy labels. …
Federated bandit: A gossiping approach
In this paper, we study Federated Bandit, a decentralized Multi-Armed Bandit problem with
a set of N agents, who can only communicate their local data with neighbors described by a …
a set of N agents, who can only communicate their local data with neighbors described by a …
Beyond images: Label noise transition matrix estimation for tasks with lower-quality features
The label noise transition matrix, denoting the transition probabilities from clean labels to
noisy labels, is crucial for designing statistically robust solutions. Existing estimators for noise …
noisy labels, is crucial for designing statistically robust solutions. Existing estimators for noise …
[HTML][HTML] Risk of bladder cancer in patients with diabetes mellitus: an updated meta-analysis of 36 observational studies
Background Increasing evidence suggests that a history of diabetes mellitus (DM) may be
associated with an increased risk of bladder cancer. We performed a systematic review with …
associated with an increased risk of bladder cancer. We performed a systematic review with …