4 ms·
I like this post, but perhaps it'd be more accurate to say that expertise is a necessary but not sufficient condition to teach a subject? I do find that the ver
by krosaen 10y ago
I like this post, but perhaps it'd be more accurate to say that expertise is a necessary but not sufficient condition to teach a subject? I do find that the very best teachers have mastery of the material.
On a related note, when I posted this notebook about the sigmoid function to HN
http://karlrosaen.com/ml/notebooks/logistic-regression-why-sigmoid/ http://karlrosaen.com/ml/notebooks/logistic-regression-why-s...
it was upvoted to the front page (presumably by non-experts like me) but derided in the comments by some experts, I think in part because it seemed so obvious to not bare explaining. So I think a challenge of having expertise is to remember what it's like to not understand something and take the beginner on the journey somehow, as the article gets at. And from this perspective a non-expert or aspiring expert can be in a sweet spot of just having learned the material him/her self. But it can be dangerous if you end up with the blind leading the blind.
- thaumasiotes 10y agoMaybe this came up in the earlier comments, but... why are you calling w_0 + w_1·x_1 + w_2·x_2 + ... + w_n·x_n a "linear combination" of the x values? It isn't. (Nor is the constant term really a "weight".)
- krosaen 10y agoHow is is w1x1 + w2x2 +... wn*xn not a linear combination? Or is your objection that, upon adding w0 it is no longer a linear combination of x1,...xn? In that case would it be more accurate to say, "a linear combination of xs plus a constant"?
- thaumasiotes 10y agoCorrect, the constant term makes the function nonlinear. A linear function must necessarily satisfy f(0) = 0 (or, I guess, f(0) = infinity).