Comparison of Pattern Recognition Techniques for Sample Classification Using Elemental Composition: Applications for ICP-AES

Abstract

Pattern recognition is very important for many aspects of data analysis and robotic control. Three pattern recognition techniques were examined—k-Nearest Neighbors, Bayesian analysis, and the C4.5 inductive learning algorithm. Their abilities to classify 71 different reference materials were compared. Each training and test example consisted of 79 different elemental concentrations. Different data sets were generated with relative standard deviations of 1, 3, 5, 10, 30, 100, and 500%. Each data set consisted of 2000 examples. These sets were used in both the training stages and in the test stages. It was found that C4.5's inductive learning algorithm had a higher classification accuracy than either Bayesian or k-Nearest Neighbors techniques, especially when large amounts of noise were present in the systems.

Keywords

Pattern recognition Induction ICP-AES

Get full access to this article

View all access options for this article.

References

Webb

D. P.

and Hamier

, and Salin

E. D.

, Trends Anal. Chem. 13, 44 (1994).

Webb

D. P.

and Salin

E. D.

, Intelligent Instruments and Computers 5, 185 (1992).

Sharif

M. A.

and Illman

D. L

, and Kowalski

B. R.

, Chemometrics (John Wiley and Sons, New York, 1986).

Duda

R. O.

and Hart

P. E.

, Pattern Classification and Scene Analysis (John Wiley and Sons, Toronto, 1973).

Quinlan

J. R.

, C4.5 Algorithms for Machine Learning (Morgan Kaufmann Publishers, San Mateo, 1993).

Salin

E. D.

and Winston

P. H.

, Anal. Chem. 64, 49A (1992).

Glick

and Hieftje

, Appl. Spectrosc. 45, 1706 (1991).

Naylorm

N. T.

Balintfy

J. L.

and Burdick

D. S

, and Chu

, Computer Simulation Techniques (John Wiley and Sons, New York, 1967).