PCA + LVQ1 - Test 10CV

Dataset			PCA/LVQ1
	vec	feat	acc.	std.	#P	std.	#K	std.
Appendicities	106	7	82.18	7.43	1.00	0.00	3.10	1.37
Australian	690	14	86.08	4.07	1.00	0.00	2.00	0.00
BitSymetry15	100	15	53.00	14.87	2.00	0.77	4.30	2.33
BitSymetry20	100	20	58.00	9.80	1.90	0.83	4.20	1.25
Breast Cancer W. (nomis)	683	9	96.19	2.10	1.00	0.00	2.00	0.00
Czerniak (trs)	250	14	70.77	8.98	1.70	0.64	4.30	0.64
Glass	214	9	58.92	7.30	1.40	0.66	4.00	0.45
Heart	270	13	80.74	9.04	1.00	0.00	2.00	0.00
Ionosphere	200	34	75.52	8.47	1.10	0.30	3.50	0.81
Iris	150	4	94.67	4.00	1.00	0.00	3.00	0.00
L. Breast (nomis)	277	9	74.03	6.91	1.00	0.00	2.10	0.30
Led500	500	7	45.51	8.60	1.50	0.67	8.20	1.08
Parity 10	1024	10	97.56	0.78	1.00	0.00	7.50	0.67
Parity 8	256	8	96.88	3.81	1.00	0.00	6.70	0.46
Voting	435	16	90.57	5.52	1.30	0.46	2.50	0.92
Wine	178	13	94.90	4.71	2.00	0.00	3.00	0.00

Network parameters

 Method            lvq+pca
 Initialization    5
 LVQ learnign rate 0.010000
 Attraction force  0.050000
 Max. projections  5
 K range           2 10 20 
 Attraction step   10
 Precision (eps)   0.030000

Dataset			QPC/LVQ1						SVM		MLP		1-NN
	vec	feat	acc.	std.	#P	std.	# K	std.	acc.	std.	acc.	std.	acc.	std.
Appendicities	106	7	85.70	1.89	1.03	0.05	2.96	0.63
Australian	690	14	86.09	0.77	1.00	0.00	2.00	0.00
Brest	683	9	97.16	0.19	1.00	0.00	2.00	0.00
Breast (Wisconsin)	683	9	97.22	0.16	1.00	0.00	2.00	0.00
Czerniak-trs	250	14	72.36	2.40	1.51	0.17	4.17	0.19
Glass	214	9	58.02	2.49	1.12	0.06	4.78	0.29
Heart	270	13	81.48	0.76	1.06	0.07	2.00	0.00
Ionosphere	200	34	78.40	1.81	1.12	0.08	3.92	0.26
Iris	150	4

— Marek 2009/03/17 15:58

PCA + LVQ1 - Test 10CV

Ustawienia

10x10CV