what type of machine learning should i implement for this case

I'm still newbie in machine learning and i need a algorithm that can study linear functions it doesn't have to be a function as i have x and y coordinates and i can feed it that,

what it should do is look at a certain points and determine if the line leading to it is straight or not calculate how many points etc... and return a value from 0 to 1 describing the probability this is event a if its not then it returns 0

for example this should return 1

while this should return 0

i looked at classifiers and i noticed its pretty much calculating the probability of both events starting from a dataset is that what ai is all about? and does deep learning work the same way?

also the features in naive bayes are not relational,in my dataset some values do have a relation and others don't if values relates to each other while they shouldn't the result will be incorrect and its impossible to fix this in the dataset

is there a way to write custom rules to classify with the sklearn module? also how would i do this exactly? and what type of ai should i use?