search

zackchase / mxnet-the-straight-dope

An interactive book on deep learning. Much easy, so MXNet. Wow. [Straight Dope is growing up] ---> Much of this content has been incorporated into the new Dive into Deep Learning Book available at https://d2l.ai/.
https://d2l.ai/
Apache License 2.0
2.56k stars 726 forks source link

Error in introduction section of "The Streight Dope" #228

Open dfisk opened 6 years ago

dfisk commented 6 years ago

Greetings, I am not sure if this is the correct place or format to log this comment, so please advise. I am just beginning to consume your notebook on deep learning and am eager to both learn and contribute. Great work! In the Introduction / Models you give an example of a loss function based on predicting the stock price of AMZN. This is fine, however there is a comment about being long or short on the stock which actually has nothing to do with the accuracy in predicting the stock price; that you may have a long or short position on the stock is irrelevant to the accuracy of the model in predicting the stock price. Kind regards, Dave

dfisk commented 6 years ago

Also, the picture labeled "Assume that you find this beautiful mushroom in your backyard:" is of a cat. A later comment is made that "as any botanist would tell us, the above actually is a death cap". No, it's cat.

zackchase commented 6 years ago

Lol there's a mystery here. The right picture shows in GitHub but the wrong picture shows in the html render.

dfisk commented 6 years ago

:-)

dfisk commented 6 years ago

Also, there is an error / typo in stating the max probability in this example: notebook: Basic probability theory section: Basic probability theory example: In [71]: totals / 1000

Out[71]: [ 0.168 0.15800001 0.15099999 0.155 0.18799999 0.18000001] <NDArray 6 @cpu(0)>

The text states: "As you can see, the lowest estimated probability for any of the numbers is about .15 and the highest estimated probability is 1.86. "

The above " highest estimated probability is ... " should read 0.188 (not 1.86)

zackchase commented 6 years ago

Ah yeah. Failed to set random seed, so you get different values every time it's run. I'll fix it. Thanks.

On Sep 22, 2017 9:39 AM, "Dave Fisk" notifications@github.com wrote:

Also, there is an error / typo in stating the max probability in this example: notebook: Basic probability theory section: Basic probability theory example: In [71]: totals / 1000

Out[71]: [ 0.168 0.15800001 0.15099999 0.155 0.18799999 0.18000001] <NDArray 6 @cpu https://github.com/cpu(0)>

The text states: "As you can see, the lowest estimated probability for any of the numbers is about .15 and the highest estimated probability is 1.86. "

The above " highest estimated probability is ... " should read 0.188 (not 1.86)

— You are receiving this because you commented. Reply to this email directly, view it on GitHub https://github.com/zackchase/mxnet-the-straight-dope/issues/228#issuecomment-331497931, or mute the thread https://github.com/notifications/unsubscribe-auth/ACR4zjwOFh_L7i82XiiPBMwz7ZtCRj-zks5sk-K1gaJpZM4PcqVA .

mbrookes commented 6 years ago

Regarding the mushroom cat issue, in the introduction the cats / dogs example has two death cap mushrooms and a box of matches where a cat should be!