2.1.4 Pattern Detector NN

← 2.1 Core NNs

xxxx

TOC

• 1 Output
• 2 PY scripts
• 4 Code with detailed comments.

For details see #607_2.1_core_NNs_.docx.

1 Output

# detection only if [1,1,1] starts at position 3 or later
if start >= 3:
y = 1
else:
y = 0

    #print(t, "pattern_score=", round(out, 4), "detected=", int(out > 0.5))
    print(t, "pattern_score=", f"{out:.8f}", "detected=", int(out > 0.5))

epoch=0 loss=0.657148
epoch=100 loss=0.239576
epoch=200 loss=0.178924
epoch=300 loss=0.142774
epoch=400 loss=0.117426
epoch=500 loss=0.103747

Tests:
[0, 0, 0, 1, 1, 1, 0, 0, 0, 0] pattern_score= 0.87180638 detected= 1
[0, 1, 0, 0, 1, 0, 0, 1, 0, 0] pattern_score= 0.00000197 detected= 0
[1, 1, 1, 0, 0, 0, 0, 0, 0, 0] pattern_score= 0.00000000 detected= 0
[0, 0, 0, 0, 0, 0, 1, 1, 1, 0] pattern_score= 0.98744279 detected= 1

2 PY scripts

• xxxxxxxxxxxx

4 Code with detailed comments

xxx

output

Extremely important concept

text2

26.0605 (v1 26.0605)