Given a dataset, (1,+), (7, - ), (2, +), (6, -), (5, +), (9, -), (11, +) You are supposed to find a threshold function that minimizes the error in the given dataset. Threshold functions look like this: f( x | a,b ) = sign(a-x).b where a is a real number and b is in {1,-1}. How many possible values should you consider to solve this problem? What is the value of a and b in the case that minimizes the error? What is the minimum error? Show how you found it.
This is a Machine Learning question :
Given a dataset,
(1,+), (7, - ), (2, +), (6, -), (5, +), (9, -), (11, +)
You are supposed to find a threshold function that minimizes the error in the given dataset. Threshold functions look like this:
f( x | a,b ) = sign(a-x).b where a is a real number and b is in {1,-1}.
How many possible values should you consider to solve this problem? What is the value of a and b in the case that minimizes the error? What is the minimum error? Show how you found it.
Note: Please first sort the dataset, and use the mid-point between two adjacent points as the threshold value to identify the misclassified points. Explain in detail how you have identified the misclassified points using the threshold value and the values of b.
Step by step
Solved in 3 steps
