mamp

12 ms·

91.
by mamp 10y ago
Yes, although in directed acyclic graphs the 'cycles' manifest as multiple paths to a node.
92.
by mamp 10y ago
It's linear in singly connected networks, not for multiply connected graphs. However NP-hard is the worst case as I mentioned (p288 Koller & Friedman)
93.
by mamp 10y ago
A heuristic guide...If you have more data than knowledge about the domain, and you want to do classification/prediction, then NNs are a good. PGMs are good if you have knowledge that is important to encode; you need modularity i.e. you
94.
by mamp 10y ago
In theory, but we didn't have enough data or computing power for machines to compete with human performance (also algorithmic improvements). I feel that the disruption isn't going to be completely new compared to past changes but
95.
by mamp 10y ago
Or lasers...
96.
by mamp 10y ago
For a post on HN the tone was ok but not great, if one of my team spoke like that in a meeting I would commend them on their technical knowledge but remind them that their effectiveness and influence depends more on tone than technical skil
97.
by mamp 10y ago
Based on what is being written about Uber's culture it's not obvious he wouldn't have been offered job if this had been known. Now it's clearly a liability.
98.
by mamp 10y ago
Looks like a great toolkit. I'm going to convert some of my PEG.js grammars over to test it out. It would be good to have an online page like https://pegjs.org/online . It allows people to start playing without all the
99.
by mamp 10y ago
MobX is great to get going quickly and it has good performance. However, in my experience, with a team Redux is great because it enforces a standard approach and clear representations. Well worth the boilerplate overhead.
100.
by mamp 10y ago
Apple did not pull Bose apps from the App Store when they pulled their products from the Apple Stores during their dispute. Hasn't happened.
101.
by mamp 10y ago
I haven't been able to find out what the battery life is with 32 Gb, but the T460 is 40% thicker than the 13" MBP due to battery, it doesn't have a Retina display and it has a slower intel graphics. I don't think there i
102.
by mamp 10y ago
Agreed, the chatbot facade slows it down, it's a simple decision tree that could be handled with a few checkboxes and conditional display.
103.
by mamp 10y ago
LawBot: Were you caused physical harm? hmm, 2 comments in and I'm confused...
104.
by mamp 10y ago
React + Flow is great also.
105.
by mamp 10y ago
Really good summary of some of the problems. There is a deeper issue which is that unless your developers understand React best practices really, really well Relay provides opportunities to screw up in a pretty big way. It goes against the
106.
by mamp 10y ago
I think you are confusing a few issues regarding 'success'. As pointed out Mycin performed well, and since de Dombal's work in the late 60's we knew that computers could perform better than experts in specific clinical d
107.
by mamp 10y ago
Interesting, but note that the Hippocratic Oath doesn't work so well with preventable errors in health care the 3rd leading cause of death in the US https://www.washingtonpost.com/news/to-your-health/wp/2
108.
by mamp 10y ago
The title is a bit misleading, the performance was memory/GC related and the specific in-memory cache that didn't work well were .Net data structures that didn't solve memory problems. He didn't try Redis or Memcache but
109.
by mamp 10y ago
Some lessons, based on the fact that as you get older your testosterone levels decrease: - Based on your testosterone levels and genetics there is a max amount of muscle you can put on per year (without drugs) so exercise and eating over t
110.
by mamp 10y ago
I didn't pick up much crap at all. I thought it was well done, and the author clearly mentioned what he preferred vs the science. I've been weightlifting for about 3 years, no hormones, I'm old (about 50), I'm a doctor a
111.
by mamp 10y ago
It depends on (a) genetics, and (b) hormone supplements. Natural body building is therefore more dependent on genetic variation and slower. There was a recent 'study' (most research in this area is pretty bad, small sample sizes,
112.
by mamp 10y ago
I think he is referring to BizTalk.
113.
by mamp 10y ago
Same here. I have some lenses that create wonderful bokeh (e.g. 85mm f/1.2). The 7 Plus bokeh effect looks like a blur, but who knows they could train a conv net to make it more bokeh like. The big advantage of the 7 Plus to me is that
114.
by mamp 10y ago
Same happened with me. I then moved to flow-type (hint, install it via 'sudo npm install -g flow-bin'), and linter-flow for atom. It's a bit temperamental but the type inference means you don't have to be quite as explic
115.
by mamp 10y ago
I've been using flow type which has made the transition from F# tolerable, so none of my assertions are type checks but still a lot of null checks because of no language support of option types.
116.
by mamp 11y ago
Our extended family have been using Path for some years ( https://path.com ). It's private so even young children can post, no ads, very slick app. It doesn't support albums but posting pictures is easy. Not sure about t
117.
by mamp 11y ago
I wonder how many LOC are for dealing with emission tests...
118.
by mamp 11y ago
A Bayesian network will update probabilities based on evidence - You can use data to learn conditional probabilities and an algorithm will update the joint probability distribution. However to suggest actions or further diagnostic tests you
119.
by mamp 11y ago
You could get away with much less memory using Redis hashes with a few more lines of code: http://redis.io/topics/memory-optimization .
120.
by mamp 11y ago
My first thought was they did it because they didn't want to rename 'Swift' to 'Drake'.

More ›