madenine

13 ms·

61.
by madenine 9y ago
I imagine getting samples/training sets that don't have some hefty source of bias is really tough in medicine, and probably leads to over-fitting.
62.
by madenine 9y ago
Its gotten to the point where the first thing I do when reading a paper is skim for their training data. If its not from a public/standard/replicable data source my bullshit meter hops up a couple steps. "We improved on the c
63.
by madenine 9y ago
Yikes. At ~60/u/mo for enterprise email & collaboration I would drop the "Replace Microsoft Outlook" line... unless you want people to start looking at what else they get beyond Outlook/Skype for Business from M
64.
by madenine 9y ago
Every time I see an article talking about how we're going to destroy the economy by turning fast food workers into kiosks and replacing teamsters with self-driving trucks... I just think how much easier it is to replace the average whi
65.
by madenine 9y ago
I'll second that. For all the attention DL/ANNs get... there's still a lot of legwork going on out there using linear models, basic trees, etc. IIRC this years kaggle survey ranked Logistic Regression as the #1 most used mode
66.
by madenine 9y ago
I think they look for modifications of the photo as well so a cropping/tilt/text overlay/etc would still be flagged. Their model might require per-processing beyond what they can or want to deploy on-device.
67.
by madenine 9y ago
Unforunately, the rise of streaming content services has kinda killed off the need for file sharing for most people. Back in the day of mytunes on a school network, I had HDs full of mp3s. I actively browsed friends (and complete strangers)
68.
by madenine 9y ago
"...pilfer Knotel’s proprietary information and trade secrets" I really want to know what kind of trade secrets a co-working space uses to its advantage
69.
by madenine 9y ago
I did a data science program with another org, and while I had an extremely positive experience and got a great job out of it - I do think the sales pitch of these programs can be very misleading, especially to those coming in with less exp
70.
by madenine 9y ago
We've started doing a lot of work with PowerBI - and I think "surprisingly good" is apt. Its got its share of flaws and issues, but for the price I think it beats everything else on the market - especially for existing Office
71.
by madenine 9y ago
https://cdn2.hubspot.net/hubfs/2172371/Q1%202017%20Gartner.p... Gartner report on commercial BI/Analytics might be useful to you
72.
by madenine 9y ago
"Machine Learning Engineer" is a title we're going to see more and more of (and we're already seeing a lot). Its one thing to know the math and theory to design, train, and tune the algorithm your company needs. But impl
73.
by madenine 9y ago
I was under the impression "Watson" was just a brand-face for IBM consulting services. You're not buying thousands consulting hours and some big data services... you're buying Watson
74.
by madenine 9y ago
If a bad actor gets your data in a breach, the article posits that only person that has to fall victim to any level of social engineering is the customer service rep of your cell provider or some other service you use. So while you might be
75.
by madenine 9y ago
Its not free, but it could be less not free... there was a time you paid for new versions of OSX.
76.
by madenine 9y ago
There are plenty of products that do this. I do some work with Power BI, and a pretty common set up is a bunch of SaaS connections (Google Analytics, Mailchimp, Salesforce, Zendesk, Stripe being the most common in my engagements), plus a da
77.
by madenine 9y ago
The maintainers! I know a couple people connected to this group - heard great things about their 2nd conference last month. The premise is great. From Russel's article on Aeon: "We organised a conference to bring the work of the m
78.
by madenine 9y ago
The article seemed to be mostly about adult-onset ADHD, which the author asserts is rare or non-existent - it would be noticed and diagnosed as a child/teenager, or evaluation would reveal that it had been ignored/untreated. I&#x2
79.
by madenine 9y ago
My experience is about the same. Video conferencing/screen share is a huge part of how my company works - and SfB is extremely reliable for us. It 'just works', if it didn't we would switch to something else immediately.
80.
by madenine 9y ago
Seems like sklearn pipelines with a more generalized use case + additional helpful features for ETL. Very interested
81.
by madenine 9y ago
you can supply a map, ie: gdf = df.groupby('col1').agg({'col2': np.mean, 'col3':np.std,'col4': lamba x: np.mean((x) / np.std(x))}) once you've got your grouped dataframe, go nuts gdf['s
82.
by madenine 10y ago
Probably should've separated/clarified my last two sentences; I agree that its already happening, and that the system will blow up on its own eventually. My questions are: Can we do something about to reverse it? If not can we do
83.
by madenine 10y ago
Its going to be a really rough bubble. There's no one source to point the finger at, but we've got ourselves in a spiral where: -More people are going to school -University costs have skyrocketed (mostly due to expanded administra
84.
by madenine 10y ago
Ignoring the comma issue, it still seems poorly worded. "The canning, processing, preserving, freezing, drying, marketing, storing, packing for shipment or distribution of:" Lets group related activities based on what workers migh
85.
by madenine 10y ago
That's usual context where I see it. Often ends up with csv's where you have columns of the data people are interested in and one column containing all the metadata bundled up in JSON as a string. Its not great to work with.
86.
by madenine 10y ago
I feel like the places where CV always proves most useful are places where either we need more eyes than would be financially viable, or we need eyes in places its hard/unsafe/expensive to have them. The tipping point on training+
87.
by madenine 10y ago
for what its worth, many of the recent competitions have had a heavy computer vision component - an area of great interest in AI research.
88.
by madenine 10y ago
From my understanding, it wasn't so much the rainfall/flow as it was the damage to the spillway. If the spillway doesn't experience structural failure, there's no problem. But it did, so flow through the spillway had to
89.
by madenine 10y ago
In the sense that there's a big, public, blowup about their culture? I think we're there.
90.
by madenine 10y ago
Its really situational. Every student at my school was given a laptop (2005, but that practice started years earlier) - but we were in a special situation being a boarding school. Every student having a laptop meant that teachers could reas

More ›