Table 2 The ranges of variations of those six training variables (v1, v2, v3, v4, v5, and v6), including the upper maximum and lower minimum limits, and MSEs regarding the entire dataset, including before and after unit conversion, are listed for those three data set #1, #2, and #3.

From: Toward fully automated UED operation using two-stage machine learning model

Variable

Formula

Description

Unit

Set 1 (10 k)

Set 2 (20 k)

Set 3 (100 k)

Summary

Min

Max

MSE

Min

Max

MSE

Min

Max

MSE

Min

Max

MSE/target

v1

xoff

x-jitter

μm

0

5

0.0731/0.3655 μm

0

5

0.0766/0.3830 μm

0

5

0.0807/0.4035 μm

0

5

0.38 μm/0.50 μm

v2

yoff

y-jitter

μm

0

5

0.0738/0.3690 μm

0

5

0.0707/0.3535 μm

0

5

0.0791/0.3955 μm

0

5

0.35 μm/0.50 μm

v3

σr

spot size

μm

30

220

0.1479/28.101 μm

30

200

0.1343/22.831 μm

30

200

0.1762/29.951 μm

30

200

22.8 μm/30 μm

v4

∆VRF/ VRF

RF amplitude jitter

 

− 0.015

0.015

0.0595/0.002

− 0.01

0.01

0.1650/0.003

− 0.025

0.025

0.1124/0.006

− 0.01

0.01

0.003/0.002

v5

θgun

RF phase

degree

5

30

0.1651/4.128°

5

60

0.0677/3.724°

5

40

0.1682/5.880°

5

60

3.7°/3°

v6

I/Imax

Solenoid factor

 

0

0.4

0.0307/3.07%

0

0.4

0.0269/2.69%

0

0.4

0.02912/2.91%

0

0.4

2.7%/3%

  1. The optimal ranges of variations, corresponding MSEs and targeted precisions of those six machine parameters are listed in the summary.