Table 2 Tabulated predictions for the dual-phase sample.

From: Efficient few-shot machine learning for classification of EBSD patterns

 

221

223

225

227

229

230

Scratch

984

0

5822

10

72

12

Transfer

1501

0

5342

0

56

1

Difference

517

0

(480)

(10)

(16)

(11)

  1. The number of patterns classified to each space group by the respective model. The difference is calculated by subtracting the number predicted by the model trained from scratch from the number predicted by the few-shot transfer learning model. Parentheses denotes the transfer learning model predicted fewer EBSPs belonging to the respective class. From left to right, the six space groups are \(Pm\overline{3}m\), \(Pm\overline{3}n\),\(Fm\overline{3}m\),\(Fd\overline{3}m\),\(Im\overline{3}m\), and \(Ia\overline{3}d\).
Back to article page