Presentation is loading. Please wait.

Presentation is loading. Please wait.

AI Practice 05 / 07 Sang-Woo Lee. 1.Usage of SVM and Decision Tree in Weka 2.Amplification about Final Project Spec 3.SVM – State of the Art in Classification.

Similar presentations


Presentation on theme: "AI Practice 05 / 07 Sang-Woo Lee. 1.Usage of SVM and Decision Tree in Weka 2.Amplification about Final Project Spec 3.SVM – State of the Art in Classification."— Presentation transcript:

1 AI Practice 05 / 07 Sang-Woo Lee

2 1.Usage of SVM and Decision Tree in Weka 2.Amplification about Final Project Spec 3.SVM – State of the Art in Classification 4.Commentary for Results on Mid-Term Project 5.Useful Technique for Final Project 6.Decision Tree © 2013, SNU CSE Biointelligence Lab., Contents 2

3 USAGE OF SVM & DECISION TREES IN WEKA

4 Neural Networks  MLP (Multilayer Perceptron)  In Weka, Classifiers-functions-MultilayerPerceptron 4© 2013, SNU CSE Biointelligence Lab.,

5 Support Vector Machines  SMO (sequential minimal optimization) for training SVM  In Weka, classifiers-functions-SMO 5© 2013, SNU CSE Biointelligence Lab.,

6 Decision Trees  J48 (Java implementation of C4.5)  In Weka, classifiers-trees-J48 6© 2013, SNU CSE Biointelligence Lab.,

7 7 Some Notes on the Parameter Setting  Parameter Setting = Car Tuning  need much experience or many times of trial  you may get worse results if you are unlucky  Multilayer Perceptron (MLP)  Main parameters for learning: hiddenLayers, learningRate, momentum, trainingTime (epoch), seed  SMO (SVM)  Main parameters: c (complexity parameter), kernel, kernel parameters  J48  Main parameters: unpruned, numFolds, minNumObj  Many parameters are for controlling the size of the result tree, i.e. confidenceFactor, pruning © 2013, SNU CSE Biointelligence Lab.,

8 8 Using SVM in Weka

9 © 2013, SNU CSE Biointelligence Lab., Using DT in Weka

10 SPEC OF FINAL PROJECT

11 Final Project (Mandatory) Extension experiments of midterm project Methods (2 mandatory algorithm) –Must Use Weka MLP (used in midterm) SVM or Decision Tree (mandatory) If you just use 2 mandatory algorithm (MLP and one more) and make reasonable result, you get default point.

12 Final Project (Optional) Methods (1 more optional algorithm) –Must Use Weka One arbitrary algorithm in weka Optional issue –Proper arbitrary algorithm for data –Proper preprocessing for data –Dataset used in midterm project If you are interested in this project, you could try optional issue and you get bonus point.

13 Final Report Final report and presentation –Submit report –About: Design your problem –Include Introduction Problem definition Dataset Preprocess (optional) Methods Experimental results and comparison Discussion Due date: PM 23:59 Submit by

14  Must Preprocess be done with weka?  No. You can use anything for preprocess  But, there is some preprocess tool in weka.  Isn’t it preferred to use Extension of Mid-term Dataset?  It is more preferred to use Extension of Mid-term Dataset.  If you do more whatever interesting for project, you will get more point!  But, it is ok just to satisfy mandatory one. © 2013, SNU CSE Biointelligence Lab., Q&A

15  C.-W. Hsu, C.-C. Chang, C.-J. Lin, “A Practical Guide to Support Vector Classification” © 2013, SNU CSE Biointelligence Lab., Some Good Materials


Download ppt "AI Practice 05 / 07 Sang-Woo Lee. 1.Usage of SVM and Decision Tree in Weka 2.Amplification about Final Project Spec 3.SVM – State of the Art in Classification."

Similar presentations


Ads by Google