Recognition of Gurmukhi Handwritten City Names Using Deep Learning and Cloud Computing

Table 2

Results obtained with a LR of 0.001 by employing Adam optimizer.

Epoch	Training loss	Validation loss	Validation accuracy (%)	Validation precision	Validation recall
1	0.49	0.08	98.5	0.98	0.97
2	0.19	0.07	98.3	0.98	0.97
3	0.13	0.05	99.2	0.99	0.99
4	0.12	0.04	99.1	0.99	0.98
5	0.09	0.02	99.3	0.99	0.98
6	0.09	0.01	99.5	0.99	0.99
7	0.09	0.03	99.6	0.99	0.98
8	0.09	0.02	99.6	0.99	0.99
9	0.09	0.02	99.8	0.99	0.99
10	0.07	0.01	99.8	0.99	0.99