Example Applications of Rough Sets Theory A

Example Applications of Rough Sets Theory – A Survey Christopher Chretien Laurentian University Sudbury, Ontario Canada October 2002

Introduction l Research on the application of Rough Sets Theory l Discovering possible areas of application l Further understanding of Rough Sets Theory usage

References l Lixiang Shen, Francis E. H. Tay, Liangsheng Qu and Yudi Shen (2000), Fault Diagnosis using Rough Sets Theory , Computers in Industry, vol. 43, Issue 1, 1 August 2000, pp. 61 -72. , URL: www. geocities. com/roughset/Fault_diagnosis_using_rough_sets_theory. pdf l Israel E. Chen-Jimenez, Andrew Kornecki, Janusz Zalewski, Software Safety Analysis Using Rough Sets, URL: http: //www-ece. engr. ucf. edu/~jza/classes/6885/rough. ps l Francis E. H. Tay and Lixiang Shen (2002), Economic and Financial Prediction using Rough Sets Model , European Journal of Operational Research 141, pp. 643 -661, URL: http: //www. geocities. com/roughset/EJOR. pdf l Pawan Lingras (2001), Unsupervised Rough Set Classification Using GAs Journal of Intelligent Information Systems, 16, 215– 228, found on: Cite. Seer, URL: http: //citeseer. nj. nec. com/cs l Rapp, S. , Jessen, M. and Dogil, G. (1994). Using Rough Sets Theory to Predict German Word Stress. in: Nebel, B. and Dreschler-Fischer, L. (Eds. ) KI-94: Advances in Artificial Intelligence, Lecture Notes in Artificial Intelligence 861, Springer-Verlag, URL: www. ims. uni-stuttgart. de/~rapp/ki 94 full. ps

Fault Diagnosis using Rough Sets Theory l Diagnosis of a valve fault for a multicylinder diesel engine l Rough Sets Theory is used to analyze the decision table composed of attributes extracted from the vibration signals

Fault Diagnosis using Rough Sets Theory l 4 states are studied among the signal characteristics Normal state l Intake valve clearance is too small l Intake valve clearance is too large l Exhaust valve clearance is too large l

Fault Diagnosis using Rough Sets Theory l 3 sampling points selected to collect vibration signals 1 st cylinder head l 2 nd cylinder head l centre of the piston stroke on the surface of the cylinder block l

Fault Diagnosis using Rough Sets Theory

Fault Diagnosis using Rough Sets Theory

Fault Diagnosis using Rough Sets Theory

Fault Diagnosis using Rough Sets Theory l 6 attributes Frequency domain attributes: IF, CG l Time domain attributes: IT, σ, Dx, α 4 l l 18 attributes for decision table l 1 decision attribute with 4 possible values based on states

Software Safety Analysis using Rough Sets l Investigates the safety aspects of computer software in safety-critical applications l Assessment of software safety using qualitative evaluations

Software Safety Analysis using Rough Sets l Use of checklists to collect data on software quality l Waterfall model Project Planning l Specification of requirements l Design l Implementation and integration l Verification and validation l Operation and maintenance l

Software Safety Analysis using Rough Sets

Software Safety Analysis using Rough Sets

Software Safety Analysis using Rough Sets l 8 student teams developing safetyrelated software Device control over the internet l Elevator controller l Air traffic control system l System satellite control system l

Software Safety Analysis using Rough Sets l 150 questions about the first 5 phases of the waterfall model l Overall safety level for 6 of the 8 projects was around 60%

Economic and Financial Prediction using Rough Sets Model l Applications of Rough Sets model in economic and financial prediction l Emphasis on main areas of business failure prediction, database marketing and financial investment

Economic and Financial Prediction using Rough Sets Model l Business l failure prediction ETEVA l Database Marketing l Financial Investment l TSE

Economic and Financial Prediction using Rough Sets Model

Economic and Financial Prediction using Rough Sets Model

Using Rough Set Theory to Predict German Word Stress l Prediction of German word stress by extracting symbolic rules from sample data l Symbolic rules are induced with a machine learning approach based on Rough Sets Theory

Using Rough Set Theory to Predict German Word Stress l Variable Precision Rough Sets Model An elementary class belongs to RβX iff a (100% - β) majority of it’s elements belongs to X l An elementary class does not belong to URβX iff a (100% - β) majority of its elements does not belong to X l

Using Rough Set Theory to Predict German Word Stress l Corpus Monomorphemic words l At least 2 non-schwa syllables l Nouns l 242 words l

Using Rough Set Theory to Predict German Word Stress l Attributes: Typ, Onset, Hoeche, Laenge, Spannung, Coda l 36 attributes in total l Attributes aligned ‘from right to left’ l Decision attribute with possible values of final, penult and antepenult

Using Rough Set Theory to Predict German Word Stress l 1 st l Stress assignment operates from right to left l 2 nd l experiment Estimate predictive accuracy l 3 rd l experiment Remove length information

Unsupervised Rough Set Classification using GAs l Rough Set classification using Genetic Algorithms l Highway classification based on predominant usage

Unsupervised Rough Set Classification using GAs l Applications of GAs Job shop scheduling l Training neural nets l Image feature extraction l Image feature identification l

Unsupervised Rough Set Classification using GAs

Unsupervised Rough Set Classification using GAs

Unsupervised Rough Set Classification using GAs

Unsupervised Rough Set Classification using GAs

Unsupervised Rough Set Classification using GAs l Rough Set classification scheme 1. 2. 3. Both uh and uk are in the same lower approximation A(Xi). Object uh is in a lower approximation and uk is in the corresponding upper approximation UA(Xi) Both uh and uk are in the same upper approximation

Unsupervised Rough Set Classification using GAs l Total error of rough set classification is the weighted sum of these errors

Unsupervised Rough Set Classification using GAs l Rough classification of highways PTC sites l Roads classified on the basis of trip purposes and trip length characteristics l Classes: commuter, business, long distance and recreational highways l Traffic patterns: hourly, daily, monthly l

Unsupervised Rough Set Classification using GAs l Experiment 264 monthly traffic patterns on Alberta highways (1987 -1991) l Rough genome consisted of 264 genes l Classes: commuter/business, long distance, recreational l

Conclusion l Triggering a better understanding of Rough Sets Theory l Opening eyes to different fields of application