Positive Example

Positive examples are increasingly central to machine learning research, particularly in scenarios with limited labeled data or high labeling costs. Current research focuses on leveraging positive examples to improve model accuracy and efficiency in various tasks, including semi-supervised learning, active learning, and PU learning, often employing contrastive learning, decision trees, and deep neural networks. These advancements are significant because they enable more efficient and robust machine learning in domains where negative examples are scarce, expensive, or difficult to obtain, with applications ranging from medical image analysis to product recommendation systems.

Papers

August 3, 2024

Safe Semi-Supervised Contrastive Learning Using In-Distribution Data as Positive Examples
Min Gu Kwak, Hyungu Kahng, Seoung Bum Kim
Contrastive Learning Loss Function Semi Supervised Self Supervised Contrastive Learning Positive Example

April 22, 2024

Ungeneralizable Examples
Jingwen Ye, Xinchao Wang
Return Conditioned Supervised Learning Unlearnable Datasets Positive Example

March 2, 2024

Improve Cost Efficiency of Active Learning over Noisy Dataset
Zan-Kai Chong, Hiroyuki Ohsaki, Bryan Ng
Active Learning Binary Classification Real World Noisy Datasets Partial Label Positive Example

January 30, 2024

Customizing Language Model Responses with Contrastive In-Context Learning
Xiang Gao, Kamalika Das
Large Language Model Contrastive Learning Contrastive Example Shot Prompting Negative Example Positive Example

November 27, 2023

Only Positive Cases: 5-fold High-order Attention Interaction Model for Skin Segmentation Derived Classification
Renkai Wu, Yinghao Liu, Pengchen Liang, Qing Chang
Positive Example Skin Segmentation High Order Attention Interaction

June 18, 2023

ProMIL: Probabilistic Multiple Instance Learning for Medical Imaging
Łukasz Struski, Dawid Rymarczyk, Arkadiusz Lewicki, Robert Sabiniewicz, Jacek Tabor, Bartosz Zieliński
Medical Imaging Multiple Instance Learning Weakly Supervised Instance Learning Instance Query Positive Example

March 14, 2023

Positive Unlabeled Learning Selected Not At Random (PULSNAR): class proportion estimation when the SCAR assumption does not hold
Praveen Kumar, Christophe G. Lambert
Positive Unlabeled Learning Random Participation Semi Supervised Classification Pu Learning Class Distribution Estimation Positive Example Scar Assumption

February 18, 2023

Data-Efficient Contrastive Self-supervised Learning: Most Beneficial Examples for Supervised Learning Contribute the Least
Siddharth Joshi, Baharan Mirzasoleiman
Contrastive Learning Self Supervised Supervised Learning Self Supervised Contrastive Learning High Quality Representation Positive Example

October 4, 2022

Automatic Generation of Product Concepts from Positive Examples, with an Application to Music Streaming
Kshitij Goyal, Wannes Meert, Hendrik Blockeel, Elia Van Wolputte, Koen Vanderstraeten, Wouter Pijpops, Kurt Jaspers
Application Proficiency Decision Tree Concept Identification Automatic Generation Positive Example

September 6, 2022

Learning Interpretable Temporal Properties from Positive Examples Only
Rajarshi Roy, Jean-Raphaël Gaglione, Nasim Baharisangari, Daniel Neider, Zhe Xu, Ufuk Topcu
Black Box Temporal Logic Interpretable Model Deterministic Finite Automaton Positive Example

Positive Example

Papers

Safe Semi-Supervised Contrastive Learning Using In-Distribution Data as Positive Examples

Ungeneralizable Examples

Improve Cost Efficiency of Active Learning over Noisy Dataset

Customizing Language Model Responses with Contrastive In-Context Learning

Only Positive Cases: 5-fold High-order Attention Interaction Model for Skin Segmentation Derived Classification

ProMIL: Probabilistic Multiple Instance Learning for Medical Imaging

Positive Unlabeled Learning Selected Not At Random (PULSNAR): class proportion estimation when the SCAR assumption does not hold

Data-Efficient Contrastive Self-supervised Learning: Most Beneficial Examples for Supervised Learning Contribute the Least

Automatic Generation of Product Concepts from Positive Examples, with an Application to Music Streaming

Learning Interpretable Temporal Properties from Positive Examples Only