DS toolbox - topic models

If you’re not primarily working with NLP you may not have been paying attention to topic modeling and word embeddings. In this post I intend to convince you that you should. ###Topic models Topic models are a set of models in NLP that discover common themes in a collection of documents. You give it a list of texts and it comes up with a bunch of topics and maps every document to a topic or a mixture of topics. …

Read on →

Lead scoring without negative examples

How do you train a binary classifier when you have only positive-labeled training examples? Impossible? Maybe. But sometimes you can do something just as good. Let’s start from the beginning… Lead generation Everyone and their mum in the b2b sector is trying to use machine learning for lead generation. For our purposes lead-gen means selecting from the millions of active companies in your country the …

Read on →

test post

This is my first post. pls work pls This is code span (hopefully) def fibo(n): if n < 2: return 1 else: return fibo(n - 1) + fibo(n - 2)

Read on →