October 25, 2016
Distributed RepresentationIn basic NLP tasks, each word is treated as discrete atomic unit. The word vector is filled with 0s and a single 1 meaning word apprearance. It’s easy to see that this one-hot representation is very sparse whose dimensionality is as large as the vocabulary size. It can be a big problem in real applications.People come up with ideas of distributed word feature representations which describes different aspects of the word and each word is associated with a point in the...
August 8, 2016
I used docker at work a couple days ago, so here to give a short overview on what is docker and how to use it.What is Docker?Docker is a hot containerization platform which wraps all the dependencies in an execution system and is easy to pack and ship as you go. What makes Docker so popular? Comparing to virtual machines, Docker is more portable and efficient. In Docker, operating systems and kernels are abstracted and shared, applications and everything else (e.g. code, runtime, system too...
March 6, 2016
Regular expression, sometimes abbreviated to Regex, is a sequence of characters that describes a search pattern in text. As a simple example, you want to list all the PNG images in a folder and the regular expression would be as easy as .+\\.png. Here . represent any character and + means 1 or more of them. We’ll discuss other marks in detail.Regular expression syntaxSpecial pattern characters have special meaning in regular expression, just like . and +. They represent a category of characte...
A data scientist, foodie,
♥ snowboarding and travelling.
