Comparison of machine learning and traditional classifiers in glaucoma diagnosis
Tóm tắt
Glaucoma is a progressive optic neuropathy with characteristic structural changes in the optic nerve head reflected in the visual field. The visual-field sensitivity test is commonly used in a clinical setting to evaluate glaucoma. Standard automated perimetry (SAP) is a common computerized visual-field test whose output is amenable to machine learning. We compared the performance of a number of machine learning algorithms with STATPAC indexes mean deviation, pattern standard deviation, and corrected pattern standard deviation. The machine learning algorithms studied included multilayer perceptron (MLP), support vector machine (SVM), and linear (LDA) and quadratic discriminant analysis (QDA), Parzen window, mixture of Gaussian (MOG), and mixture of generalized Gaussian (MGG). MLP and SVM are classifiers that work directly on the decision boundary and fall under the discriminative paradigm. Generative classifiers, which first model the data probability density and then perform classification via Bayes' rule, usually give deeper insight into the structure of the data space. We have applied MOG, MGG, LDA, QDA, and Parzen window to the classification of glaucoma from SAP. Performance of the various classifiers was compared by the areas under their receiver operating characteristic curves and by sensitivities (true-positive rates) at chosen specificities (true-negative rates). The machine-learning-type classifiers showed improved performance over the best indexes from STATPAC. Forward-selection and backward-elimination methodology further improved the classification rate and also has the potential to reduce testing time by diminishing the number of visual-field location measurements.
Từ khóa
#Machine learning #Support vector machines #Support vector machine classification #Optical receivers #Optical sensors #Machine learning algorithms #Linear discriminant analysis #Magnetic heads #Automatic testing #Multilayer perceptronsTài liệu tham khảo
box, 1973, Bayesian Inference in Statistical Analysis
lee, 0, the generalized gaussian mixture model using ica, Proc Int Workshop Independent Component Analysis (ICA 00), 239
titterington, 1985, Statistical Analysis of Finite Mixture Distributions
10.1093/bioinformatics/16.9.799
10.1093/bioinformatics/17.4.349
10.1093/bioinformatics/16.10.906
dempster, 1977, maximum likelihood from incomplete data via the em algorithm, J Roy Statist Soc Ser B, 39, 1
10.1109/72.317728
10.1016/0031-3203(92)90090-6
mclachlan, 1988, Mixture Models Inference and Applications to Clustering
10.2307/2289995
platt, 1999, probabilistic outputs for support vector machines and comparisons to regularized likelihood methods, Advances in Large Margin Classifiers
mclachlan, 1992, Discriminant Analysis and Statistical Pattern Recognition
jebara, 1999, maximum conditional likelihood via bound maximization and the cem algorithm, Advances in Neural Information Processing Systems 11, 494
schölkopf, 1996, incorporating invariances in support vector learning machines, Lecture Notes in Computer Science, 1112, 47, 10.1007/3-540-61510-5_12
jaakkola, 1999, exploiting generative models in discriminative classifiers, Advances in Neural Information Processing Systems 11, 487
10.1145/288627.288651
jaakkola, 2000, maximum entropy discrimination, Advances in Neural Information Processing Systems 12, 470
burges, 1997, improving the accuracy and speed of support vector learning machine, Advances in Neural Information Processing Systems 9, 375
sample, 1990, color permetry for assessment of primary open-angle glaucoma, Investigat Ophthalmol Visual Sci, 31, 1869
10.1097/00061198-199508000-00005
shavlik, 1990, Readings in Machine Learning
10.1097/OPX.0b013e3182004c3b
10.1136/bjo.61.2.107
broomhead, 1988, multivariable functional interpolation and adptive networks, Complex Syst, 2, 321
baum, 1988, supervised learning of probability distributions by neural networks, Advances in Neural Information Processing Systems 0, 52
rumelhart, 1986, learning internal representations by back-propagating errors, Parallel Distributed Processing Explorations in the Microstructure of Cognition, 1, 318
vapnik, 1998, Statistical Learning Theory
10.1145/130385.130401
10.1109/CVPR.1997.609310
10.1023/A:1009715923555
10.1016/B978-1-55860-335-6.50023-4
10.1016/S0004-3702(97)00063-5
ng, 2002, on discriminative vs. generative classifiers: a comparison of logistic regression and naive bayes, Advances in Neural Information Processing Systems 14
10.1001/archopht.119.7.985
turpin, 2002, development of efficient threshold strategies for frequency doubling technology perimetry using computer simulation, Investigat Ophthalmol Visual Sci, 43, 322
johnson, 1997, screening for glaucomatous visual field loos with frequency-doubling perimetry, Investigat Ophthalmol Visual Sci, 38, 413
weston, 2001, feature selection for svms, Advances in Neural Information Processing Systems 13, 668
bradley, 1998, feature selection via concave minimization and support vector machines, Proc 15th Int Conf Machine Learning, 82
ruck, 1990, feature selection using a multilayer perceptron, J Neural Network Comput, 2, 40
10.1080/00401706.1974.10489231
quinlan, 1993, C4 5 Programs for Machine Learning
10.1109/NNSP.1998.710637
bishop, 1995, Neural Networks for Pattern Recognition
10.1017/CBO9780511812651
10.1007/978-1-4757-3264-1
duda, 2001, Pattern Classification
haykin, 1999, Neural Networks A Comprehensive Foundation
10.1037/h0042519
10.1016/0893-6080(91)90047-9
10.1016/0031-3203(92)90007-6
10.1016/0167-8655(93)90098-X
cox, 1961, The Algebra of Probable Inference, 10.56021/9780801869822
chauhan, 1990, the use of visual field indices in detecting changes in the visual field in glaucoma, Investigat Ophthalmol Visual Sci, 31, 512
mitchell, 1997, Machine Learning
10.1023/A:1007465528199
saul, 2002, discriminative mixture modeling, Advances in Neural Information Processing Systems 14
cox, 1989, Analysis of Binary Data
10.2307/2531595
breiman, 1993, Classification and Regression Trees
keerthi, 1999, Improvements to platt s SMO algorithm for SVM classifier design
platt, 1998, fast training of support vector machines using sequential minimal optimization, Advances in Kernel Methods Support Vector Machines, 185
10.1148/radiology.143.1.7063747
10.1148/110.1.89
10.1214/aoms/1177728190
10.1214/aoms/1177704472
levenberg, 1944, a method for the solution of certain nonlinear problems in least square, Quart J Appl Math, ii, 164, 10.1090/qam/10666
sample, 2000, visual function specific perimetry for indirect comparison of different ganglion cell populations in glaucoma, Investigat Ophthalmol Visual Sci, 41, 1783