Never heard of it.
653 karma · joined October 2, 2016
A FT on an audio file, for example, will tell you how much of each frequency is present in that file. Each component is localized in frequency. A wavelet, however, is localized in both frequency and time (due to this decaying “hat” shape). When you scan this wavelet across the audio file and record the outputs you therefore learn not only which frequencies are present, but also their locations in the file.
> His stern ethos of self-help and bootstrapping has made him a darling of the so-called intellectual dark web, and a gateway drug for countless budding right-wingers who have stumbled upon one of his lectures on YouTube. (https://newrepublic.com/article/156829/happened-jordan-peter...)
It seems too much of a coincidence that another member of "the so-called intellectual dark web" is perhaps the most recognizable proponent of the Lab Leak Theory. She also writes for The Nation, which is unapologetically leaning pretty hard to the left.
In other words, I question both her motives and authority.
A drawback, if you want to call it that, is it can be exhausting. Both on the consumption side when you have to review lots of deeply technical and nuanced things all day and be on your A game, or in having to constantly write documents. As I’ve gotten more senior it has taken the largest part of my time now. It also seems like a lazy answer from management can be “well write that up in a doc and we can review”. Feels like you can’t discuss anything without a doc sometimes. But still love it, net net.
Tl;dr being poor isn’t as good as rich and luck matters more than zero.
Would using IP address as a feature in an online fraud risk ML model be another case of “metadata”? Speaking of, I don’t know what American Express does specifically, but I have enough experience in the fraud + ML space to confidently say that many/most of the big players are using ML in several places throughout their fraud systems, even if allow/block lists are also used.
I guess that’s my larger point: these aren’t exclusive things. Metadata is of course incredibly valuable. So is Machine Learning. They also work really well together. And sometimes simple things work best.
I’m not disagreeing with the premise that people should be more aware, just explaining how I’ve seen it happen. People excited about bringing a cool thing to market often don’t have an answer to the question, “but what if a bad actor decides to use it for nefarious purposes?”
This has led to countless headaches. The skills needed to solve these challenges are very different. My more academically minded colleagues don't seem to have the intuition for seeing around this corner and understanding what it takes to build something that actually works.
This is a long way of saying: if a company is founded by and ran by such academics, I can see how this would be a recipe for disaster when it came to actually building and shipping products.
As a person that’s built and deployed a lot of ML solutions in a related space and brought a product to market meant to help others do the same, we consistently ran into two main camps: 1) people that had a long history of rule writing and were unreasonably closed off to the idea that ML could help and 2) people that thought ML was magic and could do things it most certainly cannot.
I understand the sentiment of being skeptical in a world of outrageous claims from “fake it till you make it” startups, but the idea that most ML is not better than rule based systems is just wrong. You need both and they solve slightly different problems.
I know this because I've been on that journey. But there's no reason to expect some department head that's never been exposed to DS to know this. They just copy/paste some other company's job req. If you're more junior, here are my tips:
- If it's a "new DS team" that supports a variety of teams: beware. Bolt-on DS doesn't work well, as it's really hard to build a meaningful solution that's not deeply integrated.
- If it's an old company or in a conservative industry: beware. There are likely to be data silos and difficult ownership models that make it nearly impossible to get and join the data you need.
- If it's a small company: beware. You're likely going to need a broad set of knowledge that's won with several years of experience to be able to build end-to-end solutions that are integrated into the rest of the tech stack.
- If it's not an engineering-driven culture: beware. DS will often be used to provide evidence to someone else whose already made up their mind and pretend they're being data-driven, and you'll be the disrespected nerd that's expected to do what it takes to deliver the answer they want. Most companies claim to be "data-driven", few are, and even fewer understand data-driven isn't always possible or desirable.
Industry is still trying to figure out how to use ML and are still learning that it's not as easy as hiring someone that knows about all the algorithms, but rather it takes deep technological changes to data infrastructure to enable the datasets that can then be used by the ML experts. But you don't have to be the person that helps them figure this out the hard way (i.e. by being paid to not accomplish much due to problems outside of your control). Better to find a place with a healthy data science team that can help you learn and contribute. They exist.