Convolutional Autoencoder-Based Flaw Detection for Steel Wire Ropes

. 2020 Nov 18;20(22):6612. doi: 10.3390/s20226612

Algorithm 1. Isolation tree construction.

Isolation tree:

i T r e e (X, e, l)

Inputs:X- input data, e - current tree height, l - height limit
Output: an iTree
1: if

e \geq l

| X | \leq 1

then
2: return

e x N o d e {S i z e \leftarrow | X |}

3: else
4: let

Q

be a list of attributes in X
5: randomly select an attribute

q \in Q

6: randomly draw a split value

p

from

m a x

and

m i n

value of

q

X_{l} \leftarrow f i l t e r (X, q < p)

X_{r} \leftarrow f i l t e r (X, q \geq p)

9: return

i n N o d e {L e f t \leftarrow i T r e e (X_{l}, e + 1, l)

10:

R i g h t \leftarrow i T r e e (X_{r}, e + 1, l)

11:

S p l i t A t t \leftarrow q

12:

S p l i t V a l u e \leftarrow p}

13: end if