Community-Based Hierarchical Positive-Unlabeled (PU) Model Fusion for Chronic Disease Prediction (2023-09-06T00:00:00.000000Z)

TL;DR

This work proposes a novel approach for binary decision-making, which hierarchically builds community-based PU models and then aggregates their deliverables, and demonstrates the superior performance of PUtree as well as its variants on two benchmarks and a new diabetes-prediction dataset.

Abstract

Positive-Unlabeled (PU) Learning is a challenge presented by binary classification problems where there is an abundance of unlabeled data along with a small number of positive data instances, which can be used to address chronic disease screening problem. State-of-the-art PU learning methods have resulted in the development of various risk estimators, yet they neglect the differences among distinct populations. To address this issue, we present a novel Positive-Unlabeled Learning Tree (PUtree) algorithm. PUtree is designed to take into account communities such as different age or income brackets, in tasks of chronic disease prediction. We propose a novel approach for binary decision-making, which hierarchically builds community-based PU models and then aggregates their deliverables. Our method can explicate each PU model on the tree for the optimized non-leaf PU node splitting. Furthermore, a mask-recovery data augmentation strategy enables sufficient training of the model in individual communities. Additionally, the proposed approach includes an adversarial PU risk estimator to capture hierarchical PU-relationships, and a model fusion network that integrates data from each tree path, resulting in robust binary classification results. We demonstrate the superior performance of PUtree as well as its variants on two benchmarks and a new diabetes-prediction dataset.

Authors

Changlong Sun

5 papers

Xiaozhong Liu

2 papers

Yang Wu

1 papers

TL;DR

Abstract

Authors

References49 items

Positive-Unlabeled Learning using Random Forests via Recursive Greedy Risk Minimization

Dist-PU: Positive-Unlabeled Learning from a Label Distribution Perspective

The Prevalence and Treatment of Diabetes in China From 2013 to 2018.

Adaptive multi-task positive-unlabeled learning for joint prediction of multiple chronic diseases using online shopping behaviors

PURE: Positive-Unlabeled Recommendation with Generative Adversarial Network

Positive-Unlabeled Learning from Imbalanced Data

Predictive Adversarial Learning from Positive and Unlabeled Data

PULNS: Positive-Unlabeled Learning with Effective Negative Sample Selector

Fusion of medical imaging and electronic health records using deep learning: a systematic review and implementation guidelines

Positive Unlabeled Learning with Class-prior Approximation

Self-PU: Self Boosted and Calibrated Positive-Unlabeled Training

FNED

MNIST-NET10: A heterogeneous deep networks fusion based on the degree of certainty to reach 0.1% error rate. Ensembles overview and proposal

Do Not Have Enough Data? Deep Learning to the Rescue!

A Variational Approach for Learning from Positive and Unlabeled Data

Effects of Milk and Dairy Product Consumption on Type 2 Diabetes: Overview of Systematic Reviews and Meta-Analyses.

Revisiting Sample Selection Approach to Positive-Unlabeled Learning: Turning Unlabeled Data into Positive rather than Negative

Positive and Unlabeled Learning for Detecting Software Functional Clones with Adversarial Training

Global Model Interpretation Via Recursive Partitioning

Estimating the Class Prior in Positive and Unlabeled Data Through Decision Tree Induction

Data Augmentation by Pairing Samples for Images Classification

Generative Adversarial Positive-Unlabelled Learning

Positive-Unlabeled Learning with Non-Negative Risk Estimator

Gated Multimodal Units for Information Fusion

“Why Should I Trust You?”: Explaining the Predictions of Any Classifier

UNSW-NB15: a comprehensive data set for network intrusion detection systems (UNSW-NB15 network data set)

Improving Neural Machine Translation Models with Monolingual Data

Interpretable classifiers using rules and Bayesian analysis: Building a better stroke prediction model

Convex Formulation for Learning from Positive and Unlabeled Data

Distilling the Knowledge in a Neural Network

Supersparse linear integer models for optimized medical scoring systems

Analysis of Learning from Positive and Unlabeled Data

On the Properties of Neural Machine Translation: Encoder–Decoder Approaches

Dairy Products Consumption and Risk of Type 2 Diabetes: Systematic Review and Dose-Response Meta-Analysis

How diet modification challenges are magnified in vulnerable or marginalized people with diabetes and heart disease: a systematic review and qualitative meta-synthesis.

Do clinical standards for diabetes care address excess risk for hypoglycemia in vulnerable patients? A systematic review.

Multi-output least-squares support vector regression machines

Enhanced ocean temperature forecast skills through 3‐D super‐ensemble multi‐model fusion

Movie forecast Guru: A Web-based DSS for Hollywood managers

PEBL: Web page classification without negative examples

Building text classifiers using positive and unlabeled examples

On the privacy preserving properties of random data perturbation techniques

Partially Supervised Classification of Text Documents

Active learning using adaptive resampling

PAC Learning from Positive Statistical Queries

BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding

Language Models are Unsupervised Multitask Learners

A Study on NSL-KDD Dataset for Intrusion Detection System Based on Classification Algorithms

Random Forests

Field of Study

Journal Information

Name

Venue Information

Name

Type

URL

Alternate Names