1 Negative selection algorithms: from the thymus to V-detector Dissertation defense Zhou Ji Major professor: Prof. Dasgupta Advisory committee: Dr. Lin,

Slides:



Advertisements
Similar presentations
Applicability Issues of the (Real-valued) Negative Selection Algorithms Zhou Ji, Dipankar Dasgupta The University of Memphis GECCO 2006: July 11, 2006.
Advertisements

Augmented Negative Selection Algorithm with Variable-Coverage Detectors Zhou Ji, Zhou Ji, St. Jude Childrens Research Hospital Dipankar Dasgupta, Dipankar.
10/14/20051 Dissertation Proposal Negative selection algorithms: from the thymus to V-detector Zhou Ji, advised by Prof. Dasgupta.
V-detector: a real-valued negative selection algorithm Zhou Ji St. Jude Childrens Research Hospital.
Analysis of Dental Images using Artificial Immune Systems Zhou Ji 1, Dipankar Dasgupta 1, Zhiling Yang 2 & Hongmei Teng 1 1: The University of Memphis.
Estimating the detector coverage in a negative selection algorithm Zhou Ji St. Jude Childrens Research Hospital Dipankar Dasgupta The University of Memphis.
Negative Selection Algorithms at GECCO /22/2005.
V-Detector: A Negative Selection Algorithm Zhou Ji, advised by Prof. Dasgupta Computer Science Research Day The University of Memphis March 25, 2005.
Real-valued negative selection algorithms Zhou Ji
Applications of one-class classification
Relevant characteristics extraction from semantically unstructured data PhD title : Data mining in unstructured data Daniel I. MORARIU, MSc PhD Supervisor:
Biometrics & Security Tutorial 9. 1 (a) What is palmprint and palmprint authentication? (P10: 9-10)
My name is Dustin Boswell and I will be presenting: Ensemble Methods in Machine Learning by Thomas G. Dietterich Oregon State University, Corvallis, Oregon.
An Introduction to Artificial Intelligence By Dr Brad Morantz Viral Immunology Center Georgia State University.
Computer Vision Lecture 18: Object Recognition II
For Wednesday Read chapter 19, sections 1-3 No homework.
Comparison of Data Mining Algorithms on Bioinformatics Dataset Melissa K. Carroll Advisor: Sung-Hyuk Cha March 4, 2003.
1 BY: Nazanin Asadi Zohre Molaei Isfahan University of Technology.
Chessmen Position Recognition Using Artificial Neural Networks Jun Hou Dec. 8, 2003.
POSTER TEMPLATE BY: Multi-Sensor Health Diagnosis Using Deep Belief Network Based State Classification Prasanna Tamilselvan.
Face Recognition & Biometric Systems Support Vector Machines (part 2)
Functional Link Network. Support Vector Machines.
Anomaly Detection in Data Docent Xiao-Zhi Gao
WRSTA, 13 August, 2006 Rough Sets in Hybrid Intelligent Systems For Breast Cancer Detection By Aboul Ella Hassanien Cairo University, Faculty of Computer.
Artificial Immune Systems Andrew Watkins. Why the Immune System? Recognition –Anomaly detection –Noise tolerance Robustness Feature extraction Diversity.
Some Technical Considerations In negative selection algorithms 4/15/2005.
Content Based Image Clustering and Image Retrieval Using Multiple Instance Learning Using Multiple Instance Learning Xin Chen Advisor: Chengcui Zhang Department.
Complex Networks for Representation and Characterization of Object For CS790g Project Bingdong Li 11/9/2009.
1 Learning to Detect Objects in Images via a Sparse, Part-Based Representation S. Agarwal, A. Awan and D. Roth IEEE Transactions on Pattern Analysis and.
Evolutionary Algorithms Simon M. Lucas. The basic idea Initialise a random population of individuals repeat { evaluate select vary (e.g. mutate or crossover)
Artificial Immune Systems Our body’s immune system is a perfect example of a learning system. It is able to distinguish between good cells and potentially.
PCA Channel Student: Fangming JI u Supervisor: Professor Tom Geoden.
Smart Traveller with Visual Translator for OCR and Face Recognition LYU0203 FYP.
Generate Affy.dat file Hyb. cRNA Hybridize to Affy arrays Output as Affy.chp file Text Self Organized Maps (SOMs) Functional annotation Pathway assignment.
Multiple Object Class Detection with a Generative Model K. Mikolajczyk, B. Leibe and B. Schiele Carolina Galleguillos.
Face Processing System Presented by: Harvest Jang Group meeting Fall 2002.
Oral Defense by Sunny Tang 15 Aug 2003
Face Recognition Using Neural Networks Presented By: Hadis Mohseni Leila Taghavi Atefeh Mirsafian.
Machine Learning in Simulation-Based Analysis 1 Li-C. Wang, Malgorzata Marek-Sadowska University of California, Santa Barbara.
An Introduction to Support Vector Machines Martin Law.
By : Anas Assiri.  Introduction  fraud detection  Immune system  Artificial immune system (AIS)  AISFD  Clonal selection.
Distributed Network Intrusion Detection An Immunological Approach Steven Hofmeyr Stephanie Forrest Patrik D’haeseleer Dept. of Computer Science University.
EADS DS / SDC LTIS Page 1 7 th CNES/DLR Workshop on Information Extraction and Scene Understanding for Meter Resolution Image – 29/03/07 - Oberpfaffenhofen.
Lecture 8: 24/5/1435 Genetic Algorithms Lecturer/ Kawther Abas 363CS – Artificial Intelligence.
An Overview of Intrusion Detection Using Soft Computing Archana Sapkota Palden Lama CS591 Fall 2009.
APPLICATION OF THE METHOD AND COMBINED ALGORITHM ON THE BASIS OF IMMUNE NETWORK AND NEGATIVE SELECTION FOR IDENTIFICATION OF TURBINE ENGINE SURGING Lytvynenko.
One-class Training for Masquerade Detection Ke Wang, Sal Stolfo Columbia University Computer Science IDS Lab.
An Introduction to Support Vector Machines (M. Law)
BING: Binarized Normed Gradients for Objectness Estimation at 300fps
D. M. J. Tax and R. P. W. Duin. Presented by Mihajlo Grbovic Support Vector Data Description.
1 Classification and Feature Selection Algorithms for Multi-class CGH data Jun Liu, Sanjay Ranka, Tamer Kahveci
On Utillizing LVQ3-Type Algorithms to Enhance Prototype Reduction Schemes Sang-Woon Kim and B. John Oommen* Myongji University, Carleton University*
Supervised Machine Learning: Classification Techniques Chaleece Sandberg Chris Bradley Kyle Walsh.
Intelligent Database Systems Lab 國立雲林科技大學 National Yunlin University of Science and Technology 2005.ACM GECCO.8.Discriminating and visualizing anomalies.
Presentation By SANJOG BHATTA Student ID : July 1’ 2009.
WLD: A Robust Local Image Descriptor Jie Chen, Shiguang Shan, Chu He, Guoying Zhao, Matti Pietikäinen, Xilin Chen, Wen Gao 报告人:蒲薇榄.
Surface Defect Inspection: an Artificial Immune Approach Dr. Hong Zheng and Dr. Saeid Nahavandi School of Engineering and Technology.
Evaluation of Gender Classification Methods with Automatically Detected and Aligned Faces Speaker: Po-Kai Shen Advisor: Tsai-Rong Chang Date: 2010/6/14.
Face Detection 蔡宇軒.
High resolution product by SVM. L’Aquila experience and prospects for the validation site R. Anniballe DIET- Sapienza University of Rome.
Classifier Representation in LCS
In Search of the Optimal Set of Indicators when Classifying Histopathological Images Catalin Stoean University of Craiova, Romania
Supervised Time Series Pattern Discovery through Local Importance
Mixture of SVMs for Face Class Modeling
Machine Learning Dr. Mohamed Farouk.
Schizophrenia Classification Using
A New Approach to Track Multiple Vehicles With the Combination of Robust Detection and Two Classifiers Weidong Min , Mengdan Fan, Xiaoguang Guo, and Qing.
What I learned in the first 2 weeks
Introduction.
Immunocomputing and Artificial Immune Systems
Presentation transcript:

1 Negative selection algorithms: from the thymus to V-detector Dissertation defense Zhou Ji Major professor: Prof. Dasgupta Advisory committee: Dr. Lin, Dr. McCauley, Dr. Phan

2 Outline  Background of the research area  V-detector: a new algorithm  Experiments  Discussion on applicability and others  Conclusions

3 Background What are negative selection algorithms?

4 Related research areas Artificial Intelligence … Biology-inspired methods Neural network Evolutionary computation Artificial immune system (AIS) Negative selection algorithms Immune network Clonal selectionOther models …

5 Biological metaphor: negative selection in the thymus How T cells mature in the thymus

6 Biological metaphor: negative selection in the thymus How T cells mature in the thymus

7 Biological metaphor: negative selection in the thymus How T cells mature in the thymus  The immature T cells have diversified receptors.  Those that recognize self are eliminated.  The rest can become mature T cells.

8 Basic idea of negative selection algorithms: The problem to solve: anomaly detection or one-class classification

9 Basic idea of negative selection algorithms: Possible detectors are generated randomly.

10 Basic idea of negative selection algorithms: Those that cover self region are eliminated.

11 Components that make up a negative selection algorithm  Data and detector representation Binary (or string) representation Real-valued representation; detectors as hypersphere, or hyper-rectangle Hybrid representation  Generate/elimination mechanism Random generation + censoring Genetic algorithm Greedy algorithm or other deterministic algorithm  Matching rule Rcb (r contiguous bits) for binary representation Euclidean distance-based for real-valued representation

12 Major issues in negative selection algorithms  Number of detectors Affecting the efficiency of generation and detection  Detector coverage Affecting the accuracy of detection  Algorithm of generating detectors Linked to efficiency and quality of detector set

13 V-detector A new algorithm

14 V-detector: new development in NSA 1. Variable-sized detectors 2. Estimation of detector coverage 3. Boundary-aware interpretation of self samples 4. A generic algorithm Important features of V-detector:

15 Detectors can be just a point Detectors in their basic form (constant size) Feature 1: variable size

16 Detectors with their individual radii Detectors with maximized coverage Feature 1: variable size

17 How many detectors to generate: approach in earlier worksV-detector’s approach Feature 2: coverage estimate

18 How to estimate the coverage: Feature 2: coverage estimate A random point may be in self region in nonself region, but already covered In nonself region, not covered yet More consecutive “already covered” point  more coverage is achieved 1. An intuitive estimate; 2. hypothesis testing (Is the target coverage achieved?)

19 What does one self sample point mean? Point-wise interpretation of self samples Feature 3: boundary-aware Smaller matching thresholdLarge matching threshold

20 “The whole is more than the sum of its parts.”  Self sample could be near the boundary.  The neighboring points provide the hint. Feature 3: boundary-aware

21 How to be boundary-aware by using detectors: Feature 3: boundary-aware Point-wise interpretation Boundary-aware interpretation Large threshold Small threshold

22 V-detector as a generic algorithm  Components that can be plugged in: Data representation Distance measure Matching rule  The other three features are available for different customized variations. Feature 4: a generic algorithm

23 Example: generalized Euclidean distance Minkowski distance of order m (m-norm distance or L-m distance) Feature 4: a generic algorithm

24 Different detector shapes resulted Feature 4: a generic algorithm

25 V-detector's advantage  Efficiency: fewer detectors fast generation  Coverage confidence (reliability)  Applicable to more applications

26 Experiments V-detector in action

27 extensive experiments **  Synthetic 2-D data  Real world data Famous iris data Air pollution Biomedical data Gene expression Indian Telugu Ball bearing measurement Ball bearing measurement KDD cup data Dental image Dental image

28 2-D synthetic data Training points (1000)Test data (1000 points) and the ‘real shape’ we try to learn

29 Starting with training points, … 1000 training points 100 training points

30 Actual detectors generated Detector set based 1000 training points Detector set based 100 training points

31 Ball bearing’s structure and damage Damaged cage Raw data: measure of acceleration (time series)

32 Image-based dental diagnosis Normal occlusion Malocclusion

33 Comparison with other negative selection algorithms: iris data Training DataAlgorithmDetection RateFalse Alarm rateNumber of Detectors MeanSDMeanSDMeanSD Setosa 100% MILA * 0 NSA V-detector Setosa 50% MILA * 0 NSA V-detector Versicolor 100% MILA * 0 NSA V-detector Versicolor 50% MILA * 0 NSA V-detector Virginica 100% MILA * 0 NSA V-detector Virginica 50% MILA * 0 NSA V-detector

34 Strength of hypothesis testing ‘intersection’ shape pentagram

35 Effect of the self region’s shape stripe cross triangle ring intersection pentagram

36 Effect of the self region’s shape

37 Difference of boundary-aware interpretation (ball bearing data)

38 Comparison with SVM On disconnected 2-D self region On reduced representation of dental images

39 Discussion Whether and when are negative selection algorithm appropriate?

40 NSA’s applicability  Applicable scenario Large amount of self (normal) samples Rare or no abnormal samples  another possible usage: “negative database”  When it is not appropriate: for example, number of self samples is small.

41 Comparison with other methods  Other negative selection algorithms  SVM (Support Vector machines) One-class SVM is comparable. Kernel function is very important for SVM

42 Conclusions  Review of negative selection algorithms  V-detector: a new development High efficiency Generic algorithm  Real world application  Prospect of NSA and AIS in general

43 My publications for this dissertation  Dasgupta, Ji, Gonzalez, Artificial immune system (AIS) research in the last five years, IEEE CEC 2003  Ji, Dasgupta, Augmented negative selection algorithm with variable-coverage detectors, IEEE CEC 2004  Ji, Dasgupta, Real-valued negative selection algorithm with variable-sized detectors, GECCO 2004  Ji, Dasgupta, Estimating the detector coverage in a negative selection algorithm, GECCO 2005  Ji, A boundary-aware negative selection algorithm, ASC 2005  Ji, Dasgupta, Applicability Issues of the real-valued negative selection algorithms, GECCO 2006  Ji, Dasgupta, Analysis of Dental Images using Artificial Immune Systems, IEEE CEC 2006  Ji, Dasgupta, Revisiting negative selection algorithms, revised submission to the Evolutionary Computation Journal

44 Questions and comments? Thanks to everybody!

45 Organs in immune system