Material Identification of Particles in Space-Borne Electronic Equipments Based on Principal Component Analysis - Intelligent Computing for Sustainable Energy and Environment - page 465

0.4

0.2

0

2 4 6 8 10

Principal component

Fig. 3. Contribution rate of principal components

2

Wire
Tin
Al

1

0

−1

−2

−4 −3 −2 −1 0 1 2 3 4 5

PC1

Fig. 4. Clustering results from the PCA for different material particles

4 Classification Using SVM

In order to enhance the accuracy, the data mining method of SVM combined
with PCA were used to identify the particle material.

4.1 Basis of SVM

SVM is a machine learning method using small samples. Based on structural
risk minimization principle, SVM minimizes the empirical risk and Vapnik-
Chervonenkis (VC) dimension simultaneously. SVMs can eciently perform
non-linear classification problems by implicitly mapping their inputs into high
dimensional feature spaces using the Kernel tricks [11].
The conventional SVMs are designed for binary classification problems. In
order to train the SVM, a serial of training samples including positive and neg-
ative samples are needed. SVM aims to find a linear or nonlinear hyper-plane
with maximum margin to separate the positive and negative examples from the
training samples.