Skip to main content
Cross Validated

Return to Question

deleted 15 characters in body
Source Link
volperossa
volperossa
  • 747
  • 1
  • 5
  • 13

I read that gradient descent converge always to a local minimum while other methods as Newton's method this is not guaranteed (if the Hessian is not definite positive); but if the start point in GD is unfortunately a local maximum (and then the derivative is zero), how we can say that it converge to a (local/global) minimum? does not it stall?

I read that gradient descent converge always to a local minimum while other methods as Newton's method this is not guaranteed (if the Hessian is not definite positive); but if the start point in GD is unfortunately a local maximum (and then the derivative is zero), how we can say that it converge to a (local/global) minimum? does not it stall?

I read that gradient descent converge always to a local minimum while other methods as Newton's method this is not guaranteed (if the Hessian is not definite positive); but if the start point in GD is unfortunately a local maximum (and then the derivative is zero), how we can say that it converge to a minimum? does not it stall?

edited body
Source Link
volperossa
volperossa
  • 747
  • 1
  • 5
  • 13

I read that gradient descent converge always to a local maximumminimum while other methods as newton'sNewton's method this is not guaranteed (if the Hessian is not definite positive); but if the start point in GD is unfortunately a local maximum (and then the derivative is zero), how we can say that it converge to a (local/global) minimum? does not it stall?

I read that gradient descent converge always to a local maximum while other methods as newton's method this is not guaranteed (if the Hessian is not definite positive); but if the start point in GD is unfortunately a local maximum (and then the derivative is zero), how we can say that it converge to a (local/global) minimum? does not it stall?

I read that gradient descent converge always to a local minimum while other methods as Newton's method this is not guaranteed (if the Hessian is not definite positive); but if the start point in GD is unfortunately a local maximum (and then the derivative is zero), how we can say that it converge to a (local/global) minimum? does not it stall?

Source Link
volperossa
volperossa
  • 747
  • 1
  • 5
  • 13

gradient descent and local maximum

I read that gradient descent converge always to a local maximum while other methods as newton's method this is not guaranteed (if the Hessian is not definite positive); but if the start point in GD is unfortunately a local maximum (and then the derivative is zero), how we can say that it converge to a (local/global) minimum? does not it stall?