Simulation Visualizations

GIF

Loss vs log_2(n)

Loss vs log(n) EMPIRICAL_TRAIN

Image 1

Loss vs log(n) THEORETICAL

Image 2

Loss vs log(n) EMPIRICAL_TEST

Image 3

Loss vs log(n) 1 features

Image 4

Loss vs log(n) 2 features

Image 5

Loss vs log(n) 3 features

Image 6

Time consumption(n)

Image 1
Image 2
Image 3

Iterations vs log_2(n)

Iterations vs log(n) EMPIRICAL_TRAIN

Image 1

Iterations vs log(n) THEORETICAL

Image 2

Iterations vs log(n) EMPIRICAL_TEST

Image 3

Iterations vs log(n) 1 features

Image 4

Iterations vs log(n) 2 features

Image 5

Iterations vs log(n) 3 features

Image 6

N* Relationship Matrixes

N* theoretical

dim 1 feature(s) 2 feature(s) 3 feature(s)
1 feature(s) NaN NaN NaN
2 feature(s) NaN NaN 14.0177
3 feature(s) NaN 14.0177 NaN

N* empirical test

dim 1 feature(s) 2 feature(s) 3 feature(s)
1 feature(s) NaN NaN NaN
2 feature(s) NaN NaN 13.8083
3 feature(s) NaN 13.8083 NaN

Loss Tables

Table EMPIRICAL_TRAIN

n 1 feature(s) 2 feature(s) 3 feature(s)
2 0.000000 0.000000 0.000000
4 0.118953 0.029484 0.008906
8 0.141016 0.047835 0.020714
16 0.148484 0.059250 0.033633
32 0.154319 0.066858 0.045039
64 0.156617 0.072348 0.053324
128 0.158072 0.075086 0.057193
256 0.157725 0.076871 0.060051
512 0.158092 0.077393 0.061225
1024 0.158658 0.078044 0.062084

Table THEORETICAL

n 1 feature(s) 2 feature(s) 3 feature(s)
2 0.188731 0.149616 0.175071
4 0.185450 0.139319 0.151961
8 0.177809 0.119610 0.126663
16 0.169977 0.101823 0.100160
32 0.164718 0.091529 0.082812
64 0.161593 0.085341 0.073376
128 0.160153 0.082306 0.068356
256 0.159389 0.080505 0.065665
512 0.159020 0.079623 0.064189
1024 0.158809 0.079144 0.063494

Table EMPIRICAL_TEST

n 1 feature(s) 2 feature(s) 3 feature(s)
2 0.252575 0.172077 0.205585
4 0.205634 0.145481 0.160227
8 0.180369 0.120396 0.127551
16 0.169966 0.102151 0.100220
32 0.164836 0.091385 0.082938
64 0.161582 0.085388 0.073242
128 0.160041 0.082241 0.068144
256 0.159438 0.080571 0.065631
512 0.158843 0.079627 0.064186
1024 0.158577 0.079015 0.063448

Time Consumption Tables

Table EMPIRICAL_TRAIN

loss type time (min)
EMPIRICAL_TRAIN 0.120386
THEORETICAL 0.081022
EMPIRICAL_TEST 0.154351

Table THEORETICAL

# features time (min)
1 0.067747
2 0.059570
3 0.059700

Table EMPIRICAL_TEST

n time (min)
0 0.050629
1 0.037674
2 0.025816
3 0.017798
4 0.012664
5 0.009828
6 0.008946
7 0.010244
8 0.016076
9 0.035464

Iterations Tables

Table EMPIRICAL_TRAIN

n 1 feature(s) 2 feature(s) 3 feature(s)
2 11413 11413 11413
4 16000 16000 16000
8 11200 11200 11200
16 8000 8000 8000
32 5600 5600 5600
64 4000 4000 4000
128 2800 2800 2800
256 2000 2000 2000
512 1400 1400 1400
1024 1000 1000 1000

Table THEORETICAL

n 1 feature(s) 2 feature(s) 3 feature(s)
2 15707 22600 22600
4 9840 16000 16000
8 8064 6832 6888
16 4160 5800 5200
32 3192 4928 2996
64 2280 2480 2260
128 1540 1736 2184
256 1020 1090 1030
512 707 735 728
1024 505 505 545

Table EMPIRICAL_TEST

n 1 feature(s) 2 feature(s) 3 feature(s)
2 21922 22600 22600
4 16000 14560 14800
8 7336 9744 8232
16 4920 7680 4400
32 3108 3808 3976
64 2240 2480 2400
128 1540 1428 1484
256 1090 1150 1050
512 861 805 714
1024 725 565 560