Skip to main content
Data Science

Return to Answer

Bounty Ended with 100 reputation awarded by dnclem
added 182 characters in body
Source Link
Green Falcon
Green Falcon
  • 14.2k
  • 10
  • 58
  • 98

Consider that you are doing vector operation, change your cost function to the following:

(1 / m) * sum(((-y) .* (log(h)) - ((1 - y) .* log((1-h)))));

and your gradient to the following:

grad = (1./m) * (x' * (h - y))

Although the latter is just for precedence reassuring.

Based on the discussion in the chat, although the code calculates the cost in a wrong way, the reason the cost does not decrease is that the data is not linearly separable. Logistic regression is a simple algorithm which classifies successfully linearly separable data. Take a look at [here](https://datascience.stackexchange.com/q/21896/28175).

Consider that you are doing vector operation, change your cost function to the following:

(1 / m) * sum(((-y) .* (log(h)) - ((1 - y) .* log((1-h)))));

and your gradient to the following:

grad = (1./m) * (x' * (h - y))

Although the latter is just for precedence reassuring.

Consider that you are doing vector operation, change your cost function to the following:

(1 / m) * sum(((-y) .* (log(h)) - ((1 - y) .* log((1-h)))));

and your gradient to the following:

grad = (1./m) * (x' * (h - y))

Although the latter is just for precedence reassuring.

Based on the discussion in the chat, although the code calculates the cost in a wrong way, the reason the cost does not decrease is that the data is not linearly separable. Logistic regression is a simple algorithm which classifies successfully linearly separable data. Take a look at [here](https://datascience.stackexchange.com/q/21896/28175).
Source Link
Green Falcon
Green Falcon
  • 14.2k
  • 10
  • 58
  • 98

Consider that you are doing vector operation, change your cost function to the following:

(1 / m) * sum(((-y) .* (log(h)) - ((1 - y) .* log((1-h)))));

and your gradient to the following:

grad = (1./m) * (x' * (h - y))

Although the latter is just for precedence reassuring.