What strategy should be used to fill NaN in a pandas dataframe?

Question

I am having a csv file having a column named as Age. It is having value counts as follows.

25.0    2693
28.0    2412
30.0    2406
26.0    2391
27