24x7 Service; AnyTime; AnyWhere

Improving F-Score of the imbalance visualized pattern dataset for yield prediction robustness

Megat Mohamed Noor, Megat Norulazmi and Jusoh, Shaidah (2008) Improving F-Score of the imbalance visualized pattern dataset for yield prediction robustness. In: 21st International CODATA Conference "Scientific Information for Society - from Today to the Future" , 5 - 8 October 2008, Kyiv, Ukraine. (Unpublished)

[thumbnail of Megat_Norulazmi_Megat_Mohamed_Noor.pdf] PDF
Restricted to Registered users only

Download (3MB)


In a non closed loop manufacturing process, a prediction model of the yield outcome can be achieved by visualizing the temporal historical data pattern generated from the inspection machine, discretize to visualized data patterns, and map them into machine learning algorithm.Our previous study shows that combination of under-sampling and over sampling techniques unabel wider range of data sets where SMOTE+VDM and random under-sampling produced robust classifier performance of handling better with different batches of prediction test data.In this paper, the integration of K* entropy base similarity distance function with SMOTE, CNN+Tomek Links and the introduction of SMOTE and SMaRT (Synthetic Majority Replacement Technique)combination, has improved the classifiers F-Score robustness.

Item Type: Conference or Workshop Item (Paper)
Additional Information: CODATA (Committee on Data for Science and Technology)
Uncontrolled Keywords: Yield prediction, Predictive maintenance, Pattern visualization, Data re-sampling, Robust classifier
Subjects: Q Science > QA Mathematics
Divisions: College of Arts and Sciences
Depositing User: Mrs. Norazmilah Yaakub
Date Deposited: 18 May 2011 09:10
Last Modified: 18 May 2011 09:10

Actions (login required)

View Item View Item