(Rare) Category Detection Using Hierarchical Mean Shift Pavan Vatturi Weng-Keen Wong

Slides:



Advertisements
Similar presentations
Location Recognition Given: A query image A database of images with known locations Two types of approaches: Direct matching: directly match image features.
Advertisements

Christoph F. Eick Questions and Topics Review Dec. 10, Compare AGNES /Hierarchical clustering with K-means; what are the main differences? 2. K-means.
PARTITIONAL CLUSTERING
Machine learning continued Image source:
More on Clustering Hierarchical Clustering to be discussed in Clustering Part2 DBSCAN will be used in programming project.
Proactive Learning: Cost- Sensitive Active Learning with Multiple Imperfect Oracles Pinar Donmez and Jaime Carbonell Pinar Donmez and Jaime Carbonell Language.
Efficient Distribution Mining and Classification Yasushi Sakurai (NTT Communication Science Labs), Rosalynn Chong (University of British Columbia), Lei.
Human-Computer Interaction Human-Computer Interaction Segmentation Hanyang University Jong-Il Park.
Service Discrimination and Audit File Reduction for Effective Intrusion Detection by Fernando Godínez (ITESM) In collaboration with Dieter Hutter (DFKI)
Content Based Image Clustering and Image Retrieval Using Multiple Instance Learning Using Multiple Instance Learning Xin Chen Advisor: Chengcui Zhang Department.
Supervised learning Given training examples of inputs and corresponding outputs, produce the “correct” outputs for new inputs Two main scenarios: –Classification:
Rare Category Detection in Machine Learning Prafulla Dawadi Topics in Machine Learning.
Clustering short time series gene expression data Jason Ernst, Gerard J. Nau and Ziv Bar-Joseph BIOINFORMATICS, vol
A Study of Approaches for Object Recognition
1 DECK: Detecting Events from Web Click-through Data Ling Chen, Yiqun Hu, Wolfgang Nejdl Presented by Sebastian Föllmer.
Prénom Nom Document Analysis: Data Analysis and Clustering Prof. Rolf Ingold, University of Fribourg Master course, spring semester 2008.
Scalable Data Mining The Auton Lab, Carnegie Mellon University Brigham Anderson, Andrew Moore, Dan Pelleg, Alex Gray, Bob Nichols, Andy.
1 Efficient Algorithms for Non-Parametric Clustering With Clutter Weng-Keen Wong Andrew Moore.
A 3D Approach for Computer-Aided Liver Lesion Detection Reed Tompkins DePaul Medix Program 2008 Mentor: Kenji Suzuki, Ph.D. Special Thanks to Edmund Ng.
Segmentation by Clustering Reading: Chapter 14 (skip 14.5) Data reduction - obtain a compact representation for interesting image data in terms of a set.
Texture Reading: Chapter 9 (skip 9.4) Key issue: How do we represent texture? Topics: –Texture segmentation –Texture-based matching –Texture synthesis.
POSTER TEMPLATE BY: Note: in high dimensions, the data are sphered prior to distance matrix calculation. Three Groups Example;
Semi-Supervised Clustering Jieping Ye Department of Computer Science and Engineering Arizona State University
Pattern Recognition. Introduction. Definitions.. Recognition process. Recognition process relates input signal to the stored concepts about the object.
Clustering Ram Akella Lecture 6 February 23, & 280I University of California Berkeley Silicon Valley Center/SC.
GS 3 GS 3 : Scalable Self-configuration and Self-healing in Wireless Networks Hongwei Zhang & Anish Arora.
Birch: An efficient data clustering method for very large databases
1 Efficiently Learning the Accuracy of Labeling Sources for Selective Sampling by Pinar Donmez, Jaime Carbonell, Jeff Schneider School of Computer Science,
Graph-based consensus clustering for class discovery from gene expression data Zhiwen Yum, Hau-San Wong and Hongqiang Wang Bioinformatics, 2007.
Using Error-Correcting Codes For Text Classification Rayid Ghani Center for Automated Learning & Discovery, Carnegie Mellon University.
Methods in Medical Image Analysis Statistics of Pattern Recognition: Classification and Clustering Some content provided by Milos Hauskrecht, University.
A Geometric Framework for Unsupervised Anomaly Detection: Detecting Intrusions in Unlabeled Data Authors: Eleazar Eskin, Andrew Arnold, Michael Prerau,
Computer Vision James Hays, Brown
Employing EM and Pool-Based Active Learning for Text Classification Andrew McCallumKamal Nigam Just Research and Carnegie Mellon University.
Unsupervised Learning Reading: Chapter 8 from Introduction to Data Mining by Tan, Steinbach, and Kumar, pp , , (
Chapter 1 Introduction to Data Mining
Bayesian Sets Zoubin Ghahramani and Kathertine A. Heller NIPS 2005 Presented by Qi An Mar. 17 th, 2006.
1 Lecture 10 Clustering. 2 Preview Introduction Partitioning methods Hierarchical methods Model-based methods Density-based methods.
CSE 185 Introduction to Computer Vision Pattern Recognition 2.
Clustering Personalized Web Search Results Xuehua Shen and Hong Cheng.
EECS 274 Computer Vision Segmentation by Clustering II.
A Novel Local Patch Framework for Fixing Supervised Learning Models Yilei Wang 1, Bingzheng Wei 2, Jun Yan 2, Yang Hu 2, Zhi-Hong Deng 1, Zheng Chen 2.
(Rare) Category Detection Using Hierarchical Mean Shift Pavan Vatturi Weng-Keen Wong
Data Mining over Hidden Data Sources Tantan Liu Depart. Computer Science & Engineering Ohio State University July 23, 2012.
A Model for Learning the Semantics of Pictures V. Lavrenko, R. Manmatha, J. Jeon Center for Intelligent Information Retrieval Computer Science Department,
Data Mining Anomaly Detection Lecture Notes for Chapter 10 Introduction to Data Mining by Tan, Steinbach, Kumar © Tan,Steinbach, Kumar Introduction to.
Radial Basis Function ANN, an alternative to back propagation, uses clustering of examples in the training set.
Real-Time Tracking with Mean Shift Presented by: Qiuhua Liu May 6, 2005.
A Classification-based Approach to Question Answering in Discussion Boards Liangjie Hong, Brian D. Davison Lehigh University (SIGIR ’ 09) Speaker: Cho,
Clustering Instructor: Max Welling ICS 178 Machine Learning & Data Mining.
Cluster analysis and spike sorting
Robust Kernel Density Estimation by Scaling and Projection in Hilbert Space Presented by: Nacer Khalil.
Other Clustering Techniques
Visual Tracking by Cluster Analysis Arthur Pece Department of Computer Science University of Copenhagen
Feature Selction for SVMs J. Weston et al., NIPS 2000 오장민 (2000/01/04) Second reference : Mark A. Holl, Correlation-based Feature Selection for Machine.
Computational Biology Group. Class prediction of tumor samples Supervised Clustering Detection of Subgroups in a Class.
Clustering (1) Chapter 7. Outline Introduction Clustering Strategies The Curse of Dimensionality Hierarchical k-means.
Learning Kernel Classifiers 1. Introduction Summarized by In-Hee Lee.
Machine learning & object recognition Cordelia Schmid Jakob Verbeek.
Edge Detection using Mean Shift Smoothing
Semi-Supervised Clustering
Constrained Clustering -Semi Supervised Clustering-
Outlier Discovery/Anomaly Detection
6. Introduction to nonparametric clustering
Advanced Artificial Intelligence
Consensus Partition Liang Zheng 5.21.
CSE572, CBS572: Data Mining by H. Liu
CHAPTER 14: Information Visualization
CSE572: Data Mining by H. Liu
Presentation transcript:

(Rare) Category Detection Using Hierarchical Mean Shift Pavan Vatturi Weng-Keen Wong

1. Introduction Applications for surveillance, scientific discovery and data cleaning require anomaly detection Anomalies often identified as statistically unusual data points Many detected anomalies are simply uninteresting or correspond to known sources of noise

1. Introduction Known objects (99.9% of the data) Anomalies (0.1% of the data) Uninteresting (99% of anomalies) Interesting (1% of anomalies) Pictures from: Sloan Digital Sky Survey ( Pelleg, D. (2004). Scalable and Practical Probability Density Estimators for Scientific Anomaly Detection. PhD Thesis, Carnegie Mellon University.

1. Introduction Category Detection [Pelleg and Moore 2004]: human-in-the-loop exploratory data analysis Data Set Build Model Spot Interesting Data Points Ask User to Label Categories of Interesting Data Points Update Model with Labels

1. Introduction Data Set Build Model Spot Interesting Data Points Ask User to Label Categories of Interesting Data Points Update Model with Labels User can: Label a query data point under an existing category Or declare data point to belong to a previous undeclared category

1. Introduction Goal: present to user a single instance from each category in as few queries as possible Difficult to detect rare categories if class imbalance is severe Interested in rare categories for anomaly detection

Outline 1.Introduction 2.Related Work 3.Background 4.Methodology 5.Results 6.Conclusion / Future Work

2. Related Work Interleave [Pelleg and Moore 2004] Nearest-Neighbor-based active learning for rare-category detection for multiple classes [He and Carbonell 2008] Multiple output identification [Fine and Mansour 2006]

3. Background: Mean Shift [Fukunaga and Hostetler 1975] Reference data set Query point Center of Mass Mean shift vector (follows density gradient) Mean shift vector with kernel k

3. Background: Mean Shift [Fukunaga and Hostetler 1975] Reference data set Query point Center of Mass Convergence to cluster center

3. Background: Mean Shift Blurring Reference data set Query point Center of Mass Blurring: When query points are the same as the reference data set Progressively blurs the original data set

3. Background: Mean Shift End result of applying mean shift to a synthetic data set

4. Methodology: Overview 1.Sphere the data 2.Hierarchical Mean Shift 3.Query user

4. Methodology: Hierarchical Mean Shift Repeatedly blur data using Mean Shift with increasing bandwidth: h new = k * h old

4. Methodology: Querying the User The data point closest to the cluster center is the representative data point. Rank representative data points for querying to user according to: 1.Outlierness [Leung et al. 2000] for Cluster C i : Lifetime of C i = Log (bandwidth when cluster C i is merged with other clusters – bandwidth when cluster C i is formed)

4. Methodology: Querying the User Rank representative data points for querying to user according to: 2.Compactness + Isolation [Leung et al. 2000] for Cluster C i :

4. Methodology: Tiebreaker Ties may occur in Outlierness or Compactness/Isolation values. Highest Average Distance heuristic: choose representative data point with highest average distance from user-labeled points.

5. Results NameDimsRecordsClassesSmallest Class Largest Class Abalone %16% Shuttle %64.2% OptDigits %50% OptLetters %24% Statlog %50% Yeast %31.68% Shuttle, OptDigits, OptLetters, and Statlog were subsampled to simulate class imbalance. Data sets used in experiments

5. Results (Yeast) Category detection metric: # queries before user presented with at least one example from all categories

5. Results DatasetHMS-CIHMS- CI+HAD HMS-OutHMS- Out+HAD NNDMInterleave Abalone Shuttle OptDigits OptLetters Statlog Yeast Number of hints to discover all classes

5. Results DatasetHMS-CIHMS- CI+HAD HMS-OutNNDMInterleave Abalone Shuttle OptDigits OptLetters Statlog Yeast Area under the category detection curve

6. Conclusion / Future Work Conclusions –HMS-based methods consistently discover more categories in fewer queries than existing methods –Do not need apriori knowledge of dataset properties

6. Conclusion / Future Work Future Work Better use of user feedback Presentation of an entire cluster to the user instead of a representative data point Improved computational efficiency Theoretical analysis