Missing target variable values for a labeled model with just a time feature (timeseries)?

I am analyzing a data set with two columns (includes the target variable):

1. x – time. For example: 2020-04-02 15:30:00
2. Y- a labeled feature of two classes(1/0).

In some of the records, the target value is missing.
I think about two options:

1. Fill the target variable with the major class label (the majority class has 575 records, the minority class with 192 records, and there are 44 records with missing data)
2. Delete these records

What do you think is better? Do you have another idea?