Machine-Learning-In-Action
Machine-Learning-In-Action copied to clipboard
LogisticRegression bug?
In https://github.com/mathias-brandewinder/Machine-Learning-In-Action/blob/398677f8c739b79ff2c6f3daee4bed46f6c6fa00/MachineLearningInAction/MachineLearningInAction/LogisticRegression.fs#L96 you take len elements from a random sequence of [0..len-1]
I think this is not really correct since you might draw elements twice. That said: the overall implication is basically zero since we have mutiple rounds.
Just want to know what you are thinking about this.