Neural Networks | ML
Machine Learning Notes
Neural Networks: Representation
if there are too many features, the linear regression will be two large to be calculated. For example: suppose 100*100 pixel images (grayscale, not RGB), for term(XiXj) the features will be up to (5 * 10^7)
Cost function
- unregularized logistic regression
where
- regularized logistic regression
where
- neural network
