lack of data can be inefficiently addressed by increasing the number of training iterations. however, artificial neural networks training means ANN's weights modification , thus, using a small set of data leads to biased performance towards frequent samples of this dataset ,Moreover, ANN reveals a weak performance (sometimes break down ) against novel samples.

Dear Sharma,

The lack of training data sometimes appear in such situation. There are some methods you can try.

1. Mapping the data into new space (data transformation)

2. Data Normalization

3. Feature selection

If that method is still not enough, maybe you can change the model of ANN. Try to change activation function or add new layer.

Best Regards,

Dear Sharma,

You can see the following link:

1. http://www.opennn.net/documentation/neural_networks_basics.html

2. https://hackernoon.com/%EF%B8%8F-big-challenge-in-deep-learning-training-data-31a88b97b282