I need to come up with an algorithm that determines the probability of a check being bad based on how many good ones and bad ones were written and the date that the bad ones were written. Can anyone help?
For example, Joe writes 50 checks over a period of 5 years. 10 were bad but were written more than a year ago 20 were good but also were written more than a year ago 10 were bad and written within the past year 10 were good and written within the past year
Out of 100, what's the chance that the next check Joe writes is going to be bad keeping in mind that the older the check, the less weight it carries?