Takashi Ishida
7 papers · 2017–2024 · 4 conferences · across top CS/AI conferences
Achievements
Jump to papers ↓+3 more ↓ Show less ↑
π£ Hot Topic Early Bird π§ Keyword Pioneer π Conference Polyglot (4) π Academic Marathon (7) π Cross-Pollinator (14)
π
Interdisciplinary Bridge
πΊοΈ
Taxonomy Completionist
(10)
β
The Questioner
(2)
Conferences
ICML (3)
NIPS (2)
AISTATS (1)
ICLR (1)
Top co-authors
Keywords
weakly supervised learning
(3)
unbiased estimator
(2)
classification risk
(2)
loss function
(2)
empirical risk minimization
(1)
bregman divergence
(1)
multi-class classification
(1)
risk estimation
(1)
cross-entropy loss
(1)
training loss
(1)
conditional expectation
(1)
complementary label learning
(1)
mediated uncoupled learning
(1)
one-class classification
(1)
complementary label
(1)
double descent
(1)
positive-confidence classification
(1)
binary classification
(1)
transfer learning
(1)
supervised learning
(1)
Papers
Learning with Complementary Labels Revisited: The Selected-Completely-at-Random Setting Is More Practical
ICML 2024
Mediated Uncoupled Learning and Validation with Bregman Divergences: Loss Family with Maximal Generality
AISTATS 2023
Is the Performance of My Deep Network Too Good to Be True? A Direct Approach to Estimating the Bayes Error in Binary Classification
ICLR 2023
Do We Need Zero Training Loss After Achieving Zero Training Error?
ICML 2020
Complementary-Label Learning for Arbitrary Losses and Models
ICML 2019
Binary Classification from Positive-Confidence Data
NIPS 2018
Learning from Complementary Labels
NIPS 2017