Can someone explain what the fbeta score here signify? I am a beginner.
1 Like
Hey,
Basically it is one way to check our model’s accuracy, this is enough to know for this course, but if oy want to know more read on:
To understand this, you must understand what true positives are. When you train any model you are going to have following possiblity:
- The model predicted yes and the actual output was also yes, it is called True Positives
- The model predicted yes but the actual output was no. It is called a False Positive
- The model predicted No but the actual output was yes. It is called a False Negative
- The model predicted No and the actual output was also No, it is called True Negative
Now Precision is defined as True Positives/All Positives and recall is defined as True Positives/(True positives + False Negatives).
Basically Precision means what part of Positives you predicted are actually positive, and recall measures what amount of actual positive did you predict right.
Generally we want both of these values to be high, so to compute a model that does both very well, we use F score.
Regards,
Jayam.
2 Likes