Comparison of data reduction techniques based on SVM classifier and SVR performance

Abstract

In this work, we applied several data compression techniques to simulated data and the Turbofan engine degradation simulation data set from NASA, with the goal of comparing their performance when coupled with the Support Vector Machine (SVM) classifier and the SVM regression (SVR) predictor. We consistently attained correct rates in the neighborhood of 90% for simulated data set, with the Principal Component Analysis (PCA), Sparse Reconstruction by Separable Approximation (SpaRSA) and Partial Least Squares (PLS) having a slight edge over the other data reduction methods for data classification. We achieved 22% error rate with SRM for the Turbofan data set 1 and 40% error rate with PCA for Turbofan data set 2. Throughout the tests we have performed, PCA proved to be the best data reduction method.

Temporal Classification of Turbofan Engine Health using Elman Recurrent Network

“…The author used DBN which is also a probabilistic model. As already mentioned w could not made a fair comparison to Zhao et al [10] since the lack of methodological details. From now on, the results to be reported originated from experiments designed to closely follow the approach by Tamilselvan and Wang [9].…”

“…Misclassifications tend to occur between states HS-2 and HS-3. The Elman network misclassified HS-1 and HS-2 as HS-3 [1,2,3,6,8,10,11,12,13,14,19,20] with higher rates. However, it should be noted that the power of this network comes from its short-term memory ability.…”

“…Tamilselvan and Wang [9] achieved over 90% of accuracy, being a state of art approach on this problem. Zhao et al [10] also reports 90% of accuracy on their paper. However, there are quite a few missing details about the used methodology for a fair comparison.…”

“…The confusion matrix of Elman on OS-1 subset of data. The features are the sensors[1,2,3,6,8,10,11,12,13,14,19,20] …”

Review and Analysis of Algorithmic Approaches Developed for Prognostics on CMAPSS Dataset

“…Publication ID (T. Wang, Yu, Siegel, & Lee, 2008) 1 (Heimes, 2008) 2 (Peel, 2008) 3 (Coble & Hines, 2008) 4 (Coble, 2010) (Coble & Hines, 2011) (Siegel, 2009) 5 (Ramasso, 2009) 6 (T. Wang, 2010) 7 (Riad, Elminir, & Elattar, 2010) 8 (Abbas, 2010) 9 (Ramasso & Gouriveau, 2010) 10 (Ramasso & Gouriveau, 2013) (Sarkar, Jin, & Ray, 2011) 11 (Xue, Williams, & Qiu, 2011) 12 (Zhao, P., & Willett, 2011) 13 (El-Koujok, Gouriveau, & Zerhouni, 2011) 14 (Liao & Sun, 2011) 15 (P. 16 (Son, Fouladirad, & Barros, 2012) 17 (Richter, 2012) 18 (Sun, Zuo, Wang, & Pecht, 2012) 19 (Peng, Wang, Wang, Liu, & Peng, 2012) 20 (Hu, Youn, Wang, & Yoon, 2012) 21 (Peng, Xu, Liu, & Peng, 2012) 22 (Javed, Gouriveau, Zemouri, & Zerhouni, 2012) 23 (Serir, Ramasso, & Zerhouni, 2012) 24 25 (Yu, 2013) 26 (Ramasso, Rombaut, & Zerhouni, 2013) 27 (Liu, Gebraeel, & Shi, 2013) 28 (Son, Fouladirad, Barros, Levrat, & Iung, 2013) 29 (Xi, Jing, Wang, & Hu, 2013) 30 (Lin, Chen, & Zhou, 2013) 31 (Javed, Gouriveau, & Zerhouni, 2013) 32 (Ramasso & Denoeux, 2013) 33 (Li, Qian, & Wang, 2013) 34 (Tamilselvan & Wang, 2013) 35 (Ishibashi & Nascimento Junior, 2013) 36 (Gouriveau, Ramasso, & Zerhouni, 2013) 37 (Jianzhong, Hongfu, Haibin, & Pecht, 2010) 38…”