fish_reg

fish_reg

date: Fri, 12 Oct 2007 13:57:53
SVN revision: 146

units used: 512
train objects: 5

data properties:

	#bins
phi_y	36
y	36
x	36
z	36
id	10

parameters:

n_bins	36
reg_train_fraction	0.5
phase_b_normalised	True
KNN_k	0
class_train_fraction	0.5
class_test_fraction	1.0

Histograms

Regressions

Regression for cos(y/2) -> y

errors:

	old obj.	new obj.
mean dev.	0.04	0.10
std. dev.	0.05	0.14

LaTeX

b vector:

index	2	1	3	11	12	17
value	0.657	-0.208	-0.114	-0.062	0.053	0.047

Regression for cos(x/2) -> x

errors:

	old obj.	new obj.
mean dev.	0.03	0.08
std. dev.	0.05	0.12

LaTeX

b vector:

index	1	2	3	4	11	13
value	-0.676	-0.203	0.087	-0.083	0.048	-0.044

Regression for cos(z/2) -> z

errors:

	old obj.	new obj.
mean dev.	0.15	0.31
std. dev.	0.18	0.39

LaTeX

b vector:

index	44	9	43	50	49	14
value	0.157	0.138	0.136	0.120	-0.104	0.092

Regression for cos, sin -> phi_y

errors:

	old obj.	new obj.
mean dev.	42.61	79.78
std. dev.	58.90	96.53

LaTeX

b vector:

index	10	47	91	81	102	3
value	0.148	0.111	0.110	0.108	0.102	0.099

LaTeX

Per Object Regression for phi_y

old objects:

errors for old objects:

object	1	2	3	4	5
mean dev.	9.60	23.31	16.70	32.19	30.28
std. dev.	13.12	39.76	25.49	51.46	45.74

new objects:

errors for new objects:

object	6	7	8	9	10	mean
mean dev.	16.60	27.83	22.18	32.98	22.66	24.45
std. dev.	23.61	45.27	33.95	49.98	35.92	37.74

LaTeX

errors for all objects:

object	1	2	3	4	5	6	7	8	9	10	mean
mean dev.	9.60	23.31	16.70	32.19	30.28	16.60	27.83	22.18	32.98	22.66	23.43
std. dev.	13.12	39.76	25.49	51.46	45.74	23.61	45.27	33.95	49.98	35.92	36.43

LaTeX