4 ms·
These sorts of things are typically better offered as libraries, particularly as the training is usually specific to a corpus, or a particular context. It woul
by mattmcknight 13y ago
These sorts of things are typically better offered as libraries, particularly as the training is usually specific to a corpus, or a particular context.
It would be a nice to offer a library with a bootstrapped training set.
- phillmv 13y agoNot to mention either the sensitivity behind the data, the sheer volume behind it, or the effort involved in customizing it for a particular algorithm or input - only for it to shut down and take your data with you. Machine Learning as a Service seems Hella Neat, tho.
- LambdaAlmighty 13y agoSorry, don't understand your last sentence. It seems to contradict the paragraph before -- ML as a service seems a terrible idea for the reasons you just listed (among others). What's "Hella Neat" about that?
- phillmv 13y agoThe problem mostly stems from the vast risk you take on from making a large investment in an unstable/unproven platform vendor. Servers are relatively fungible, given ops automation; it's painful but not the end of the world if you have to migrate away. But the technology is still relatively immature in that building your own ML service in house - and having it scale, etc - is still a big pain. I would immensely prefer it if we first brought ML libraries up to a higher level of maturity - as simple as apt-get install and adding `includes ActiveLearning::Bayes` to your models. But if a client came to me tomorrow and said "there's this great Amazon API that we're thinking of using" I wouldn't consider that insane on first principles.