abhgh

19 ms·

181.
by abhgh 6y ago
For many corporate client system developments this is my experience too. Producing the actual LOCs is not the primary cost; it's requirement gathering, figuring out how to plug in the new component into an existing system etc. As an ex
182.
by abhgh 6y ago
That'd be a very useful resource. Another example is the use in impedances [1] [1] http://hyperphysics.phy-astr.gsu.edu/hbase/electric/impcom.h...
183.
by abhgh 6y ago
I have mentioned this twice before on HN, and I have been making progress: I am interested in creating "compact" (self-defined, non-standard term) machine learning models. A compact model is a smaller version of a model that in ge
184.
by abhgh 6y ago
I partly work in this space, and I can confirm that geo-fencing over a time period is the probably the most transparent and convincing (to the client) way to do this. Essentially you mark out local regions (as small as zip codes or as large
185.
by abhgh 6y ago
This is an interesting project, thank you!
186.
by abhgh 6y ago
Thank you! Yes, I wasn't thinking about the read-time support.
187.
by abhgh 6y ago
Unfortunately my school expects a PDF thesis. But you make a good point about popular alternate readers not supporting animations - maybe this is a wasted effort. Thank you! Probably better to link to notebooks or videos of the animations
188.
by abhgh 6y ago
I am considering animations for my thesis. When printing, a designated frame should be used, but inside a reader, the animation should work. I am writing my thesis in LaTeX too. Any pointers?
189.
by abhgh 6y ago
In the specific case of ML courses - many of which I have TA-ed or attended classes of, this reason does not ring true at all. Libraries for most standard algorithms are available in some form with a Python interface (or for the more statis
190.
by abhgh 6y ago
I don't think deep work means no meetings, it's avoidance of pointless meetings. The github page links to a guide which explicitly discusses "collaborative deep work".
191.
by abhgh 6y ago
I identify with the "rituals" bit a lot. My most vivid example if this is from a project, much of whose code I had co-written. It was a real-time model delivery system, which mostly used to work fine, except for a few minutes ever
192.
by abhgh 6y ago
PyMC4 on TFP was the plan, but they made a recent announcement [1] indicating those efforts would stop, and instead, they would develop PyMC3+JAX+Theano. [1] https://pymc-devs.medium.com/the-future-of-pymc3-or-theano-i...
193.
by abhgh 6y ago
I agree with this ... I coded a Thompson NFA [1] out of interest a few years ago; definitely recommended as an exercise. [1] https://swtch.com/~rsc/regexp/regexp1.html
194.
by abhgh 6y ago
I am following Tommi Jakkola in this space too. Also would recommended Ameet Talwalkar's group, esp the papers with Gregory Plumb like MAPLE [1]. [1] [PDF] https://papers.nips.cc/paper/2018/file/b495ce63e
195.
by abhgh 6y ago
I have made another comment on this page and this is indeed a problem barring specific cases where you have a "simulator" and you want to learn something applicable to some "aggregate of simulations". For ex. learning a
196.
by abhgh 6y ago
Agree with your point about anonymization, but here is a call out for generated data too: if the objective is to build a model, you might end up losing information, or worse, your model might end up modelling the assumptions in the generat
197.
by abhgh 6y ago
I agree about Perl; haven't done a lot of one liners with Ruby when I was using the language. Perl was made to be really compact to the detriment of being hard to read. I think Python strikes a nice balance, and you can compress it to
198.
by abhgh 6y ago
Yes and no. But even for cases, where the algorithm is not published, the general approach can be indicated and hence "Ask them where on Wikipedia you can read more about the approach they took" might be still valid. For ex. you c
199.
by abhgh 6y ago
Same here ... Went back to look at that image after the parent comment!
200.
by abhgh 6y ago
My primary use case has been to download talks and watch them later. I frequently used to work out of cafes (not since Feb of course) where the connection may not be reliable. Youtube-dl was very helpful in those cases.
201.
by abhgh 6y ago
I am sure this is mostly seen as a positive but this can also backfire in an interview, and later, at work (I see a sibling comment addresses this). If a person leads with saying he has written a book on a topic, and then is unable to provi
202.
by abhgh 6y ago
Very enjoyable series, and David Suchet is stellar. It is also interesting to see some of the cast who went on to greater fame like Alice Eve, Peter Baelish and Talulah Riley; there are probably more I'm unable to remember right now.
203.
by abhgh 6y ago
This is probably already well known, but if you're using Python and want to exploit SIMD instructions, one of the simpler ways is to reformulate stuff as matrix operations and let numpy handle them. The speed-up can be staggering. Tens
204.
by abhgh 6y ago
Avid note-taker here. I agree with you in that the two mediums - (1) a note taking software where you type in things (2) writing notes on paper - have different benefits and use-cases. In my case, I pretty much now know what kind goes whe
205.
by abhgh 6y ago
I can vouch for the gate with a guardhouse. Was turned away by the guard because I am not a PA resident (I live in Mtn View).
206.
by abhgh 6y ago
I think many had a similar experience with Firebug, including me. I remembering coming across it for the first time in an issue of Dr Dobb's. I read the article wide-eyed. Truly a "game changer" worthy of the term.
207.
by abhgh 6y ago
Thank you for the links! I find Adobe Color quite helpful to find a palette for visualization [1]. Its been around for years now, although I think the original website was called Adobe Kuler. For categorical colormaps, I have found the pyth
208.
by abhgh 6y ago
I recommend the book in general, and also his lectures. He has a quirky sense of humour, compares his book with Harry Potter here: http://www.inference.org.uk/mackay/itila/Potter.html
209.
by abhgh 6y ago
I was about to post that article myself; good write-up on the pitfalls of interpreting t-sne viz.
210.
by abhgh 6y ago
Do you any plans to support math symbols in annotations?

More ›