Breeding Decision Trees Using Evolutionary Techniques Papagelis Athanasios - Kalles Dimitrios Computer Technology Institute & AHEAD RM.

Slides:



Advertisements
Similar presentations
Genetic Algorithms Chapter 3. A.E. Eiben and J.E. Smith, Introduction to Evolutionary Computing Genetic Algorithms GA Quick Overview Developed: USA in.
Advertisements

Is Random Model Better? -On its accuracy and efficiency-
DECISION TREES. Decision trees  One possible representation for hypotheses.
Random Forest Predrag Radenković 3237/10
1 Machine Learning: Lecture 3 Decision Tree Learning (Based on Chapter 3 of Mitchell T.., Machine Learning, 1997)
Using Parallel Genetic Algorithm in a Predictive Job Scheduling
CPSC 502, Lecture 15Slide 1 Introduction to Artificial Intelligence (AI) Computer Science cpsc502, Lecture 15 Nov, 1, 2011 Slide credit: C. Conati, S.
1er. Escuela Red ProTIC - Tandil, de Abril, Decision Tree Learning 3.1 Introduction –Method for approximation of discrete-valued target functions.
For Friday Finish chapter 5 Program 1, Milestone 1 due.
Advancements in Genetic Programming for Data Classification Dr. Hajira Jabeen Iqra University Islamabad, Pakistan.
A new crossover technique in Genetic Programming Janet Clegg Intelligent Systems Group Electronics Department.
Tree-based methods, neutral networks
Three kinds of learning
Artificial Intelligence Genetic Algorithms and Applications of Genetic Algorithms in Compilers Prasad A. Kulkarni.
Evolutionary Computational Intelligence Lecture 8: Memetic Algorithms Ferrante Neri University of Jyväskylä.
Genetic Algorithms: A Tutorial
A Genetic Algorithms Approach to Feature Subset Selection Problem by Hasan Doğu TAŞKIRAN CS 550 – Machine Learning Workshop Department of Computer Engineering.
Breeding Decision Trees Using Evolutionary Techniques Written by Athanasios Papagelis Dimitris Kalles Presented by Alexandre Temporel.
Machine Learning Chapter 3. Decision Tree Learning
Efficient Model Selection for Support Vector Machines
Charles L. Karr Rodney Bowersox Vishnu Singh
For Wednesday No new reading Homework: –Chapter 18, exercises 3, 4, 7.
1 Paper Review for ENGG6140 Memetic Algorithms By: Jin Zeng Shaun Wang School of Engineering University of Guelph Mar. 18, 2002.
C. Benatti, 3/15/2012, Slide 1 GA/ICA Workshop Carla Benatti 3/15/2012.
Cristian Urs and Ben Riveira. Introduction The article we chose focuses on improving the performance of Genetic Algorithms by: Use of predictive models.
SOFT COMPUTING (Optimization Techniques using GA) Dr. N.Uma Maheswari Professor/CSE PSNA CET.
An Introduction to Artificial Life Lecture 4b: Informed Search and Exploration Ramin Halavati In which we see how information.
CS 484 – Artificial Intelligence1 Announcements Lab 3 due Tuesday, November 6 Homework 6 due Tuesday, November 6 Lab 4 due Thursday, November 8 Current.
For Friday No reading No homework. Program 4 Exam 2 A week from Friday Covers 10, 11, 13, 14, 18, Take home due at the exam.
GATree: Genetically Evolved Decision Trees 전자전기컴퓨터공학과 데이터베이스 연구실 G 김태종.
Genetic Algorithms Michael J. Watts
What is Genetic Programming? Genetic programming is a model of programming which uses the ideas (and some of the terminology) of biological evolution to.
Optimization Problems - Optimization: In the real world, there are many problems (e.g. Traveling Salesman Problem, Playing Chess ) that have numerous possible.
1 “Genetic Algorithms are good at taking large, potentially huge search spaces and navigating them, looking for optimal combinations of things, solutions.
Neural and Evolutionary Computing - Lecture 9 1 Evolutionary Neural Networks Design  Motivation  Evolutionary training  Evolutionary design of the architecture.
GENETIC ALGORITHM A biologically inspired model of intelligence and the principles of biological evolution are applied to find solutions to difficult problems.
Derivative Free Optimization G.Anuradha. Contents Genetic Algorithm Simulated Annealing Random search method Downhill simplex method.
Kansas State University Department of Computing and Information Sciences CIS 732: Machine Learning and Pattern Recognition Friday, 16 February 2007 William.
For Wednesday No reading Homework: –Chapter 18, exercise 6.
For Monday No new reading Homework: –Chapter 18, exercises 3 and 4.
CS 8751 ML & KDDDecision Trees1 Decision tree representation ID3 learning algorithm Entropy, Information gain Overfitting.
Genetic Algorithms ML 9 Kristie Simpson CS536: Advanced Artificial Intelligence Montana State University.
For Friday Finish chapter 6 Program 1, Milestone 1 due.
MACHINE LEARNING 10 Decision Trees. Motivation  Parametric Estimation  Assume model for class probability or regression  Estimate parameters from all.
1 Universidad de Buenos Aires Maestría en Data Mining y Knowledge Discovery Aprendizaje Automático 5-Inducción de árboles de decisión (2/2) Eduardo Poggi.
1 Decision Tree Learning Original slides by Raymond J. Mooney University of Texas at Austin.
Coevolutionary Automated Software Correction Josh Wilkerson PhD Candidate in Computer Science Missouri S&T.
Data Mining and Decision Support
D Nagesh Kumar, IIScOptimization Methods: M8L5 1 Advanced Topics in Optimization Evolutionary Algorithms for Optimization and Search.
Advanced AI – Session 6 Genetic Algorithm By: H.Nematzadeh.
Genetic Algorithms An Evolutionary Approach to Problem Solving.
An Evolutionary Algorithm for Neural Network Learning using Direct Encoding Paul Batchis Department of Computer Science Rutgers University.
Genetic Algorithm. Outline Motivation Genetic algorithms An illustrative example Hypothesis space search.
Evolutionary Computation Evolving Neural Network Topologies.
Introduction to genetic algorithm
Rule Induction for Classification Using
Who cares about implementation and precision?
Issues in Decision-Tree Learning Avoiding overfitting through pruning
Advanced Artificial Intelligence Evolutionary Search Algorithm
Introduction to Data Mining, 2nd Edition by
Genetic Algorithms: A Tutorial
Machine Learning Chapter 3. Decision Tree Learning
Machine Learning: Lecture 3
Machine Learning Chapter 3. Decision Tree Learning
GATree Genetically Evolved Decision Trees
Search.
Traveling Salesman Problem by Genetic Algorithm
Search.
Using Bayesian Network in the Construction of a Bi-level Multi-classifier. A Case Study Using Intensive Care Unit Patients Data B. Sierra, N. Serrano,
Genetic Algorithms: A Tutorial
Presentation transcript:

Breeding Decision Trees Using Evolutionary Techniques Papagelis Athanasios - Kalles Dimitrios Computer Technology Institute & AHEAD RM

Introduction We use GAs to evolve simple and accurate binary decision trees Simple genetic operators over tree structures Experiments with UCI datasets very good size competitive accuracy results Experiments with synthetic datasets Superior accuracy results

Current tree induction algorithms….. Use greedy heuristics To guide search during tree building To prune the resulting trees Fast implementations Accurate results on widely used benchmark datasets (like UCI datasets) Optimal results ? No Good for real world problems? There are not many real world datasets available for research.

More on greedy heuristics They can quickly guide us to desired solutions On the other hand they can substantially deviate from optimal WHY? They are very strict Which means that they are VERY GOOD just for a limited problem space

Why GAs should work ? GAs are not Hill climbers  Blind on complex search spaces Exhaustive searchers  Extremely expensive They are … Beam searchers  They balance between time needed and space searched Application on bigger problem space Good results for much more problems No need to tune or derive new algorithms

Another way to see it.. Biases Preference bias Characteristics of output  We should choose about it e.g small trees Procedural bias How we will search?  We should not choose about it  Unfortunately we have to: Greedy heuristics make strong hypotheses about search space GAs make weak hypotheses about search space

The real world question… Are there datasets where hill-climbing techniques are really inadequate ? e.g unnecessarily big – misguiding output Yes there are… Conditionally dependent attributes  e.g XOR Irrelevant attributes  Many solutions that use GAs as a preprocessor so as to select adequate attributes Direct genetic search can be proven more efficient for those datasets

The proposed solution Select the desired decision tree characteristics (e.g small size) Adopt a decision tree representation with appropriate genetic operators Create an appropriate fitness function Produce a representative initial population Evolve for as long as you wish!

Initialization procedure Population of minimum decision trees Simple and fast Choose a random value as test value Choose two random classes as leaves A=2 Class=1 Class=2

Genetic operators

Payoff function Balance between accuracy and size set x depending on the desired output characteristics. Small Trees ?  x near 1 Emphasis on accuracy ?  x grows big

Advanced System Characteristics Scalled payoff function (Goldberg, 1989) Alternative crossovers Evolution towards fit subtrees  Accurate subtrees had less chance to be used for crossover or mutation. Limited Error Fitness (LEF) (Gathercole & Ross, 1997) significant CPU timesavings and insignificant accuracy loses

Second Layer GA Test the effectiveness of all those components coded information about the mutation/crossover rates and different heuristics as well as a number of other optimizing parameters Most recurring results: mutation rate crossover rate 0.93 use a crowding avoidance technique Alternative crossover/mutation techniques did not produce better results than basic crossover/mutation

Search space / Induction costs 10 leaves,6 values,2 classes Search space >50,173,704,142,848 (HUGE!) Greedy feature selection O(ak) a=attributes,k=instances (Quinlan 1986) O(a 2 k 2 ) one level lookahead (Murthy and Salzberg, 1995) O(a d k d ) for d-1 levels of lookahead Proposed heuristic O(gen * k 2 * a). Extended heuristic O(gen * k * a)

How it works? An example (a) An artificial dataset with eight rules (26 possible value, three classes) First two activation-rules as below:  (15.0 %) c1  A=(a or b or t) & B=(a or h or q or x)  (14.0%) c1  B=(f or l or s or w) & C=(c or e or f or k) Huge Search Space !!!

How it works? An example (b)

Illustration of greedy heuristics problem An example dataset (XOR over A1&A2) A1A2A3Class TFTT TFFT FTFT FTTT FFFF FFFF TTTF TTFT

C4.5 result tree A1=t A2=t t f A2=f f t Totally unacceptable!!! A1=t A2=t t f A2=f f t A3=t

More experiments towards this direction NameAttrib.Class FunctionNoiseInstanc. Random Attributes Xor1 10 (A1 xor A2) or (A3 xor A4) No1006 Xor2 10 (A1 xor A2) xor (A3 xor A4) No1006 Xor3 10 (A1 xor A2) or (A3 and A4) or (A5 and A6) 10% class error 1004 Par1 10 Three attributes parity problem No1007 Par2 10 Four attributes parity problem No1006

Results for artificial datasets C4.5GATree Xor167± ±0 Xor253± ±17.32 Xor379±6.5278±8.37 Par170±24,49100±0 Par263±6.7185±7.91

Results for UCI datasets

C4.5 / OneR deficiencies Similar preference biases Accurate, small decision trees  This is acceptable Not optimized procedural biases Emphasis on accuracy (C4.5)  Not optimized tree’s size Emphasis on size (OneR)  Trivial search policy Pruning as a greedy heuristic has similar disadvantages

Future work Minimize evolution time crossover/mutation operators change the tree from a node downwards we can classify only the instances that belong to the changed-node’s subtree. But we need to maintain more node statistics Average needed re-classification

Future work (2) Choose the output class using a majority vote over the produced tree forest (experts voting) Pruning is a greedy heuristic A GA’s pruning?