Hamel Husain

…u probably should be excited about ML. There’s some seriously amazing progress happening out there. But let’s get practical first. And incrementally prove our impact as Data Scientists before reaching for the big, expensive hammer.

…own to provide return on our effort to make it so, we can build out extremely innovative solutions. But building out an extremely innovative solution to a non-problem is a recipe for some uncomfortable realignment with your management.

ULMFiT is a method that uses a pre-trained model on millions of Wikipedia pages that can be tuned in for a specific task. This tuned-in model is later used to create a classifier. This approach is impressively efficient: “with only 100 labeled examples (and giving it access to a…

Acquiring Quality Labels Is Challenging

…rmulation requires deep domain knowledge, the ability to decompose problems, and a lot of patience. The most convenient training dataset should not drive how we formulate the problem, rather it should be the other way around. This is an important first skill to becoming an effective problem solver in machine learning.

Finally we show that using the representation learnt by the discriminator we can attain competitive results to using other representation learning methods for the MNIST dataset and the insurance dataset such as a wide variety of autoencoders as shown in …

Jeremy Howard recommends developers new to git check out hub (https://github.com/github/hub7) . I don’t have personal experience with it, but I’m told that forking isn’t required and that you can…

When we see successful data products we often see expertly designed user interfaces with intelligent capabilities and most importantly, a useful output which, at the very least, is perceived by the users to solve a pertinent problem. Now if a data scientist spends their time only learning how to write and execute machine learning algorithms, then they can only be a small (albeit necessary) part of a team that leads to the success of a project that produces a valuable product. This means that data science teams that work in isolation will struggle to provide value!

Create a place to track the trash talk so you can throw it back in people’s faces later.