Classification from Positive, Unlabeled and Biased Negative Data - PowerPoint PPT Presentation

Dec 31, 2022 •367 likes •505 views

Classification from Positive, Unlabeled and Biased Negative Data Poster #180 Yu-Guan Hsieh 1 , Gang Niu 2 , Masashi Sugiyama 2,3 1 ENS Paris, France 2 RIKEN, Japan 3 The University of Tokyo, Japan Poster #180 Background and problem setup 1 / 7

Classification from Positive, Unlabeled and Biased Negative Data Poster #180 Yu-Guan Hsieh 1 , Gang Niu 2 , Masashi Sugiyama 2,3 1 ENS Paris, France 2 RIKEN, Japan 3 The University of Tokyo, Japan
Poster #180 Background and problem setup 1 / 7
Poster #180 Background and problem setup Supervised Positive (P) Negative (N) 1 / 7
Poster #180 Background and problem setup Supervised Semi-supervised Positive (P) Positive Negative (N) Negative Unlabeled (U) 1 / 7
Poster #180 Background and problem setup PUbN Supervised Semi-supervised Positive Positive (P) Positive Biased Negative (bN) Negative (N) Negative Unlabeled Unlabeled (U) 1 / 7
Poster #180 Background and problem setup PUbN Supervised Semi-supervised Positive Positive (P) Positive Biased Negative (bN) Negative (N) Negative Unlabeled Unlabeled (U) 1 / 7
Poster #180 Motivating examples Positive Samples Labeled Negative Samples Other Negative Samples ● Information retrieval, text classification, sentiment analysis ● Medical diagnosis: healthy population requesting physical exams is biased 2 / 7
Poster #180 Method: Empirical risk estimator Unbiased Estimator Risk Minimization Unbiased labeled data Empirical Risk Minimization 3 / 7
Poster #180 Method: Empirical risk estimator σ(x) = p(s=+1|x) probability of x being labeled η>0 determining how much we rely on the U data to approximate the risk 4 / 7
Poster #180 Method: Empirical risk estimator #P data #bN data #U data σ(x) = p(s=+1|x) probability of x being labeled η>0 determining how much we rely on the U data to approximate the risk 4 / 7
Poster #180 Method: Illustration Step 1 Step 2 bN P P Regarded as N U y = +1 y = -1 σ ↑ final classifier: y as label ERM: estimate σ = p(s=+1|.): s as label pseudo labeling + weight adjustment nnPU classifier (Kiryo+ NeurIPS 2017) 5 / 7
Poster #180 Estimation error bound With probability at least 1-δ #P data #bN data #U data Bias due to inexact approximation of σ 6 / 7
Poster #180 Experiments Models: ConvNet / ResNet / FCN + Training: Amsgrad Dataset P π bN ρ nnPU/nnPNU PUbN(\N) PU→PN Not given NA 5.76 ± 1.04 4.64 ± 0.62 NA MNIST 2, 4, 6, 8, 10 0.49 1, 3, 5 0.3 5.33 ± 0.97 4.05 ± 0.27 4.00 ± 0.30 9 > 5 > others 0.2 4.60 ± 0.65 3.91 ± 0.66 3.77 ± 0.31 Not given NA 12.02 ± 0.65 10.70 ± 0.57 NA Airplane, Cat, dog, horse 0.3 10.25 ± 0.38 9.71 ± 0.51 10.37 ± 0.65 CIFAR-10 automobile, 0.4 ship, truck Horse > deer 0.25 9.98 ± 0.53 9.92 ± 0.42 10.17 ± 0.35 = frog > others Not given NA 23.78 ± 1.04 21.13 ± 0.90 NA Cat, deer, dog, CIFAR-10 0.4 Bird, frog 0.2 22.00 ± 0.53 18.83 ± 0.71 19.88 ± 0.62 horse Car, truck 0.2 22.00 ± 0.74 20.19 ± 1.06 21.83 ± 1.36 Not given NA 14.67 ± 0.87 13.30 ± 0.53 NA sci. 0.21 14.69 ± 0.46 13.10 ± 0.90 13.58 ± 0.97 20 alt., comp., 0.56 Newsgroups misc., rec. talk. 0.17 14.38 ± 0.74 12.61 ± 0.75 13.76 ± 0.66 soc. > talk. > sci. 0.1 14.41 ± 0.76 12.18 ± 0.59 12.92 ± 0.51 7 / 7

Recommend

Clustering Clustering is an unsupervised classification method, i.e. unlabeled data is partitioned

Clustering Clustering is an unsupervised classification method, i.e. unlabeled data is partitioned into subsets (clusters), according to a similarity measure, such thatsimilardata is grouped into the same cluster. Unlabeled Data

442 views • 22 slides

Classification from Pairwise Similarity and Unlabeled Data Han Bao 1,2 , Gang Niu 2 , Masashi

Classification from Pairwise Similarity and Unlabeled Data Han Bao 1,2 , Gang Niu 2 , Masashi Sugiyama 2,1 1 The University of Tokyo, Japan / 2 RIKEN, Japan July 13 th , 2018 Gentle Start: Binary Classification 2 Boundary Training data

264 views • 13 slides

Cla lass Prior Shif ift and Asymmetric Error Nontawat Charoenphakdee 1,2 and Masashi Sugiyama 2,1

Positive-Unlabeled Cla lassification under Cla lass Prior Shif ift and Asymmetric Error Nontawat Charoenphakdee 1,2 and Masashi Sugiyama 2,1 The University of Tokyo 1 RIKEN AIP 2 2 Supervised binary ry classification (P (PN classification)

581 views • 24 slides

Nearest Neighbor Classification Seed classification by area and What should we compactness

Nearest Neighbor Classification Seed classification by area and What should we compactness predict for unlabeled Lecture 6 KNN and Decision Trees test points (stars)? Nearest neighbor classification: predict CS 335 label of

203 views • 3 slides

10701 Semi supervised learning Can Unlabeled Data improve supervised learning? Important

10701 Semi supervised learning Can Unlabeled Data improve supervised learning? Important question! In many cases, unlabeled data is plentiful, labeled data expensive Image classification (x=images from the web, y=image type) Text

886 views • 37 slides

Learning from Unlabeled Video Carl Vondrick Columbia University Survivor Bias of Video Data

Learning from Unlabeled Video Carl Vondrick Columbia University Survivor Bias of Video Data Large-scale Video Classification with Convolutional Neural Networks, CVPR 2014 Survivor Bias of Video Data Large-scale Video Classification with

760 views • 39 slides

Mimicking Word Embeddings using Subword RNNs Yuval Pinter, Robert Guthrie, Jacob Eisenstein

Mimicking Word Embeddings using Subword RNNs Yuval Pinter, Robert Guthrie, Jacob Eisenstein @yuvalpi Presented at EMNLP September 2017, Copenhagen The Word Embedding Pipeline Unlabeled Unlabeled corpus Unlabeled corpus Unlabeled corpus

1.27k views • 126 slides

10 Steps to Counting Unlabeled Planar Graphs: 20 Years Later Manuel Bodirsky October 2007

10 Steps to Counting Unlabeled Planar Graphs: 20 Years Later Manuel Bodirsky October 2007 Counting Unlabeled Planar Graphs A005470 Sloane Sequence A005470 (core, nice, hard): Number p ( n ) of unlabeled planar simple graphs with n nodes.

814 views • 70 slides

Sparse representation classification and positive L 1 minimization Cencheng Shen Joint Work with

Sparse representation classification and positive L 1 minimization Cencheng Shen Joint Work with Li Chen, Carey E. Priebe Applied Mathematics and Statistics Johns Hopkins University, August 5, 2014 Cencheng Shen (JHU) JSM2014 Presentation

345 views • 30 slides

Introduction to Machine Learning Evaluation: Measures for Binary Classification: ROC

Introduction to Machine Learning Evaluation: Measures for Binary Classification: ROC visualization compstat-lmu.github.io/lecture_i2ml LABELS: ROC SPACE Plot True Positive Rate and False Positive Rate: 1.00 True Class y C2 +

596 views • 17 slides

A biased history of equality in type theory Some equations are more equal than others James

A biased history of equality in type theory Some equations are more equal than others James Chapman Institute of Cybernetics, Tallinn Semantics days - Andu A typical biased history Famous work by genius 350 BC (e.g. Aristotle) time All of

396 views • 22 slides

Biased landscape in random constraint satisfaction problems Louise Budzynski LPENS, PhD with

Biased landscape in random constraint satisfaction problems Louise Budzynski LPENS, PhD with Guilhem Semerjian June 30, 2020 Table of contents 1. Introduction 2. Biased measure over the set of solutions 3. Large k asymptotics of the

348 views • 22 slides

Word2Vec Michael Collins, Columbia University Motivation We can easily collect very large

Word2Vec Michael Collins, Columbia University Motivation We can easily collect very large amounts of unlabeled text data Can we learn useful representations (e.g., word embeddings) from unlabeled data? Bigrams from Unlabeled Data

254 views • 6 slides

Extreme Event-Size Extreme Event-Size Fluctuations in Biased Fluctuations in Biased Random

Extreme Event-Size Extreme Event-Size Fluctuations in Biased Fluctuations in Biased Random Walks on Networks Random Walks on Networks Vimal Kishore Physical Research Laboratory Ahmedabad, India phy.vimal@gmail.com vimal@prl.res.in Plan of

817 views • 25 slides

Lecture 26 ANNOUNCEMENTS Homework 12 due Thursday, 12/6 OUTLINE Self-biased current sources

Lecture 26 ANNOUNCEMENTS Homework 12 due Thursday, 12/6 OUTLINE Self-biased current sources Self-biased current sources BJT MOSFET Guest lecturer Prof. Niknejad EE105 Fall 2007 Lecture 26, Slide 1 Prof. Liu, UC Berkeley

589 views • 17 slides

Becky Coffin Kingfisher plc Net Positive 2 Net Positive 3 Net Positive 4 Creating the

Achieving Net Positive through responsible sourcing Becky Coffin Kingfisher plc Net Positive 2 Net Positive 3 Net Positive 4 Creating the leader Net Positive 5 Net Positive 6 WHAT HAVE WE LEARNED ABOUT RESPONSIBLE SOURCING?

450 views • 19 slides

Learning How to Move and Where to Look from Unlabeled Video Kristen Grauman Department of

Learning How to Move and Where to Look from Unlabeled Video Kristen Grauman Department of Computer Science University of Texas at Austin Visual recognition Objects amusement park sky Activities Scenes Locations The Wicked Cedar Point

768 views • 60 slides

Positive Discipline The Solution Studio What is Positive Discipline? Positive discipline is a

Positive Discipline The Solution Studio What is Positive Discipline? Positive discipline is a key to helping children develop self-discipline, responsibility, cooperation, and problem-solving skills. It is a discipline model used by schools,

381 views • 6 slides

Linear Classification w T x i is the classifier score for the instance x i The score can be used

Linear Classification w T x i is the classifier score for the instance x i The score can be used in different ways to make a classification. Perceptron: output positive class if score is at least 0, otherwise output negative class Today:

679 views • 30 slides

Graph Classification Classification Outline Introduction, Overview Classification using

Graph Classification Classification Outline Introduction, Overview Classification using Graphs Graph classification Direct Product Kernel Predictive Toxicology example dataset Vertex classification Laplacian Kernel

605 views • 33 slides

Unlabeled Motzkin numbers Max Alekseyev Dept. Computer Science and Engineering 2013 Max

Unlabeled Motzkin numbers Max Alekseyev Dept. Computer Science and Engineering 2013 Max Alekseyev Unlabeled Motzkin numbers Catalan numbers Catalan numbers can be defined by the explicit formula: 2 n 1 (2 n )! C n = = n + 1 n !( n +

690 views • 20 slides

Biased and Unbiased Samples James J. Heckman Econ 312, Spring 2019 May 14, 2019 1 / 125

Definitions and Some Examples of Biased Samples Biased and Unbiased Samples James J. Heckman Econ 312, Spring 2019 May 14, 2019 1 / 125 Definitions and Some Examples of Biased Samples Definitions and Some Examples of Biased Samples All

1.42k views • 127 slides

Biased and Unbiased Samples James J. Heckman Econ 312, Spring 2019 May 13, 2019 1 / 125

Definitions and Some Examples of Biased Samples Biased and Unbiased Samples James J. Heckman Econ 312, Spring 2019 May 13, 2019 1 / 125 Definitions and Some Examples of Biased Samples Definitions and Some Examples of Biased Samples All

1.3k views • 127 slides

Combining Biased and Unbiased Estimators in High Dimensions Bill Strawderman Rutgers University

Combining Biased and Unbiased Estimators in High Dimensions Bill Strawderman Rutgers University (joint work with Ed Green, Rutgers University) OUTLINE : I. Introduction II. Some remarks on Shrinkage Estimators III. Combining Biased and

532 views • 24 slides