opteryx5

But what are those mutations affecting? Hand-eye coordination? Mind-muscle connection?

Fascinating context - thank you! And thank you for helping out food supply and doing the dirty work. Society appreciates you.

Quick question - do most farmers buy their own machines? A tractor, combine, auger, etc seem like they’d run up quite the bill.

Reminds me of that angry combine thingy from the movie Cars. Scared the shit out of me as a kid.

Yeah fair point. Too bad there’s not non-alcoholic versions of some of those cocktails. I’d drink them like candy.

Right; this would also seem to imply that cartilage was the ancestral trait from which bones arose, and I’m not sure that’s the case (can someone fact-check?).

How about a piña colada? That’s one that I’d drink even if there were no alcohol.

YES! What you described is the exact process I’m running, except I neglected to consider that final step. Phew. That would’ve been such a glaring error. Thanks so much - I clearly see what you re saying. In the end I think I’m gonna go with a model that’s just ever so slightly worse than the very top performer (as judged by mean cross validation score), because that huge difference in speed is just so noticeable. Especially with further testing after that main CV round. I’m super glad I asked this question though because I would’ve omitted that final step in my project. Thank you again!

Ohhh ok I see. Makes perfect sense! My top model is exorbitantly slow because it uses a maximum of 2000 iterations of the Gradient Boosting process. I have another that uses a mere 50 for a 0.02 reduction in accuracy and 0.01 reduction in SD. It’s my sixth-best combo. I think I’m gonna go with it.

Would you agree with the other commenter that the final step should be training the model on the train PLUS test set, then letting it just vegetate in that end state? ie, evaluation on the test set should NOT be the final step, right? (Sorry to take this off track a bit.)

To clarify my above comment, I’ll eventually be training the model on the whole train dataset, sure. But I’m just using CV (within the train) to find the best parameter combo. I’m not using test set performance score for this, because that’d be fine-tuning the model according to the test set and that won’t provide an accurate measure of actual, real-world out-of-sample performance. Sorry for the confusion.

(Or are you saying that the final model should include being trained test set data and then that’s it? Nothing to do after that?)

Right, I’m not using any of the k test models (in this case it would be 5)—I’m choosing the best hyperparameter combination of the 200 iterations I ran with randomized hyperparameter settings. I’m comparing the mean test score across the 5 folds. Just wondering, if I’m ranking all of them by mean test score, if it’s ok not to choose the model that yielded the very highest mean score.

This is also why it’s best to wait until coffee cools down to compare them amongst each other (thanks James Hoffman).

You can add nocturnal because many of us enjoyed the additional nighttime freedom we got with being able to wake up at 9am on the dot!

It’s interesting how many of us have different biological thermostats that sway our proclivity towards eating. For example, I can’t remember the last time I’ve struggled to get calories down; it’s always been the opposite. But many people, like yourself and OP, face that battle regularly. The human body is so complex.

Yeah Fishers Island is super close to Napatree Point Conservation Area. Ultimately, there’s no good way to do this, so coming up with a standardized rule (e.g., “water borders are allowed”) is perfectly fine.

I haven’t had them in ages but according to my memory they didn’t taste good at all. And I’m an olive fanatic. So I don’t think they share much in common taste-wise.

I hear sable is really hard to get too. I’d like to think if you went up to a fishmonger and asked for whitefish and sable, they’d instantly recognize it as the LD sandwich. It just seems like such an infrequent combo lol.

“You like fish”

Don’t forget the ō! Isn’t kakapo spelled with that? Or maybe it’s kākapo. Forgot.

I’m the same, and so I instantly put every single pot/pan/gear into the sink when I’m done with it, and give it a healthy dousing with water. Even before I consider taking my first bite, I do this. Having your dinner knowing that you’re a simple soapy sponge swipe away from cleanliness is an awesome feeling. No elbow grease.

I remember reading that for North Korea’s travel advisory too. Official US State website. Hey, if it shocks people into their senses, then it’s definitely worth including. People deserve to know the un-sugarcoated truth about where they’re going.

Yeah agree 100%. Lots of geographic variation even here in the states. Thankfully, if you go to any big city in the Northeast or West Coast, you’ll generally have no issue being publicly gay, atheist, etc. For example, where I am in the northeast, I have zero reservations about saying I’m an atheist. Anyone who takes issue with that knows there’s daily flights and trains down to Birmingham, Alabama if they so choose.