jwuphysics

Maybe I'm just an "academic in charge" but I don't understand how MOND being able to account for galaxy rotation curves means that dark matter is not real. There is missing matter in galaxy clusters. Matter is missing if we look at galaxy-galacy and cluster-galaxy strong and weak gravitational lensing. This matter appears to be non-interacting (see Bullet Cluster, El Gordo, etc). There is missing matter from the angular power spectrum of the cosmic microwave background if we only account for baryons. There is missing matter if we are to forward model gravitational collapse of large scale structure at early times. MOND doesn't explain any of these things.

Oriental Raisin Tea is a common hangover cure in East Asia. It's not clear to me whether the relevant compounds were covered in this study, but I would love to see more exploration of how well these Asian remedies work. (I'm personally skeptical of 中医 = Traditional Chinese Medicine, but am open to being convinced otherwise.)

Have you written about your experiences developing this firmware and your involvement with the Telescope Array? I would be very interested to hear some of the stories (and lessons learned)!

Based on the OP's username, surely one of the deep learning algorithms is a denoising autoencoder, right?

No. The experimentalists used lasers to condense the atoms to high densities.

I think that's fairly typical for people who aren't used breaking down images into their constituent parts. In the language of computer vision, it would be like saying that our visual cortexes don't need to know how to segment a face into detailed descriptions of eyes, a nose, a mouth, etc., in the same way as it is necessary to be able to segment a wolf from a forest scene.

But it can also be trained. See, e.g., Drawing on the Right Side of the Brain by Betty Edwards.

Here is the actual article: https://www.pnas.org/content/118/7/e2013284118

> We tested eight annual and seasonal climate variables, including temperature, precipitation, frost days, and atmospheric CO2 concentrations in a mixed-effects model framework to account for city-to-city variation. We found that mean annual temperature was the strongest predictor of these four pollen metrics (P < 0.0001 for all metrics; Fig. 2 and SI Appendix, Table S4). The full mixed-effects models explained 51–90% of the variance (i.e., conditional R^2) in pollen metrics, and mean annual temperature alone (i.e., marginal R^2) explained 14–37% of the variance in pollen metrics (SI Appendix, Figs. S4 and S5 and Table S4). Notably, while atmospheric CO2 concentrations were sometimes included in the group of the most parsimonious models, the variation explained was often quite low (e.g., annual integral R^2 marginal = 0.01).

I'm not sure what I expected, but temperature fluctuations seem like a sensible primary driver of pollen count. Other variables that I think might be interesting include wind speed and relative humidity.

Arthletic Brewing offers some really amazing Beliner Weisse beers and IPAs. I made the partial switch to NA beers over a year ago and it's been amazing. Only problem is that hoppy NA beers still trigger my GERD...

I had exactly the same reaction, and I also went to CMU for undergrad (not SCS though). However I found that it didn't translate to long-term productivity during my PhD program, when I needed to think about my career goals 5 years in the future. There I needed to focus on sustainable work ethic and working "smarter" rather than "harder" -- for example, okay I got an A in my quantum field theory class, but who cares? Other students who took easier courses but were able to start writing papers probably got ahead in the long run.

The URL points to labml.ai. Also, HN encourages matching the post title with the link title.

While I respect the publication of this idea in a research journal, I'm fairly disgruntled by how Avi Loeb is now spending so much of his efforts promoting his book on the matter (i.e. profiting). The vast majority of astrophysicists reject the light sail theory as an explanation for the anomalous acceleration that 'Oumuamua experienced, now that we have better physical models.

This would pose a different kind of problem for people with aphantasia.

I can also provide some input on how to insult different kinds of astronomers depending on the field they work in.

Cosmology: Use the word "cosmology" interchangeably with "astronomy."

Simulations: "Did you include dust?" If they say yes then ask about magnetic fields. Either way, claim that they tuned their physical hyperparameters to achieve their results.

Observations: "Did you follow up with <X telescope at different wavelength>?" Pick one that's especially competitive to request observations, such as Hubble or SOFIA or ALMA, so that if the answer is no they'll feel extra bad about their rejected Hubble/SOFIA/ALMA observing proposal.

Cheers from Charm City! I just wanted to note that your website's listing says that ML engineers will be mainly working in TensorFlow (Python) and Scala.

I'm a little tired of articles mentioning "Harvard Astronomer" to add credibility to this ridiculous theory. Nobody else seriously believes this and we all know Avi wants to sell his book. Out of my network of ~200 astronomers/physicists, I haven't heard one person support the idea that this is an alien probe.

Astronomy is also an interesting case because this law doesn't necessarily hold true. The power spectrum, or Fourier transform of the autocorrelation function, isn't monotonic for matter on large scales. For example, baryonic acoustic oscillations at early times in the Universe get "imprinted" into the cosmic structure, such that galaxies have a preferential separation between each other.

Wow, is that the auditorium in the CMU Gates building? I remember pulling all-nighter "homework parties" in there!

It doesn't appear that the Swift port for fast.ai is a high priority, but Jeremy, Sylvain, and co have rewritten the Python library from scratch and it is much more sensibly organized now. The version 2 codebase is completed developed (generated, along with documentation) from Jupyter notebooks.

I've just received mine and I must say I find it so much nicer to learn using the print copy book! It forces me to type out (rather than copy and paste) the code, and I like being able to scribble all over the book. Given the thousands of dollars of value I've gained from the courses, I have to say that spending another $60 for the book was well worth it.

I searched on my podcast app, and found an RSS feed that has the three existing episodes. Who would have known that YC has a SoundCloud account?

https://feeds.soundcloud.com/users/soundcloud:users:24682936...

That is because redder colors correspond to a cooler or lower color temperature. Colloquially, we think of blues as cool and reds as warm, but blue light is actually more energetic (or "hot") than red light.

I can't get through the payway, but based on The Verge (http://www.theverge.com/2016/1/27/10841600/starry-wireless-g...) story and comments that I've seen here, I think I'm going to have to be pessimistic on this one. As a millimeter to radio wavelength astronomer, I wish them luck transmitting signals through any sort of inclement weather or in locations of high humidity.

The factor that the Comet Lovejoy contains ethyl alcohol and a simple saccharide is not unexpected; this paper does however mark the first identification of these organic molecules in a comet. Astronomers have previously detected both of these, along with plenty of other complex organic molecules, in other regions of the Milky Way (e.g., https://en.wikipedia.org/wiki/List_of_interstellar_and_circu...).

Likewise: http://imgur.com/ux3xfL5

I've never seen this venn diagram before--thanks for bringing it up. I find that, as an academic (pursuing a Ph.D. in astrophysics) that plenty of traditional researchers are able to hack together code (many haven't ever taken a formal programming course; http://arxiv.org/abs/1507.03989) but many also misuse or can't interpret statistics (from personal experience). That puts us in the danger zone!

I think the key is to find the mathematics and statistics interesting because you want the [data] science to be meaningful. If that's a driving force, then you can learn math and statistics on your own (like the author did). Otherwise, yes--you will find yourself in the danger zone.

Oh boy. I also think that string theories (in their current state) are wrong and possibly a dead end, but I feel uneasy reading this. Besides the harmless typos [0] in the article and his questionable academic affiliation (iSETI [1], which does not pass the straight face test), his referenced publication [2] is published in a questionable journal [3] and begins with an abstract that jumps around from topic to topic while "introducing" new terminology such as "kenos" and tau, a "change in time dilation over a specific height."

[0] E.g., "Noble Prize"

[1] http://www.iseti.us/

[2] http://www.scirp.org/Journal/PaperInformation.aspx?PaperID=3...

[3] https://en.wikipedia.org/wiki/Scientific_Research_Publishing

To add to this list, the gravitational waves interpretation of BICEP2's B-mode polarization signal is incorrect.

Pretty much everyone actually thinks black holes are real. I am a physicist.

This is definitely a nice, intuitive way to introduce people to the theory of gravity. However, I was a little confused about how he tried to model dark energy.