Skip to main content

Official Journal of the Asia Oceania Geosciences Society (AOGS)

Table 5 The CPU time (seconds) against different numbers of neurons utilized by the VQTAM and VQTAM-LLE approaches

From: Simulating 2-D magnetotelluric responses using vector-quantized temporal associative memory artificial neural network-based approaches

Num. of neurons

Stopping criteria of objective functions

\(\mathbf {1.0\%}\)

\(\mathbf {0.5\%}\)

\(\mathbf {0.1\%}\)

\(\mathbf {0.0 5\%}\)

VQTAM

VQTAM-LLE

VQTAM

VQTAM-LLE

VQTAM

VQTAM-LLE

VQTAM

VQTAM-LLE

\(10 \times 10\)

5

6

5

6

5

5

5

5

\(15 \times 15\)

8

8

83

84

124

124

146

147

\(20 \times 20\)

23

23

134

134

153

154

170

171

\(25 \times 25\)

21

22

222

223

238

239

251

252

\(30 \times 30\)

118

119

338

339

387

388

401

402

\(35 \times 35\)

165

166

509

510

613

615

651

653

\(40 \times 40\)

743

745

801

803

955

957

1096

1098