Description
Representation Learning and Deep Sequential Modeling for Predicting Topics of Sentences in Job Advertisements by Clemens Westrup
Recent advances in Machine Learning, specifically in Deep Learning have led to a popularization of this discipline beyond research and even into mainstream media. Approaches using Deep Learning and the so-called area of Representation Learning are now being applied to various areas of science and engineering and an active open source developer community has evolved around them pushing the limits of these algorithms and making them widely available and applicable for everyone.
In this talk we will look at some of the latest developments using these techniques in the context of Natural Language Processing, in particular looking at the problem of text classification. A special emphasis will be placed on open source tools and libraries in Python that allow cutting edge research while allowing for fast, iterative and explorative research.
About the author: Clemens Westrup recently wrote his Master's Thesis on Machine Learning for Natural Language Processing in Aalto University's Machine Learning and Data Mining programme. Originally just visiting Finland for an exchange year from Germany he fell in love with the place and the people when finding fellow musicians and falling in love with the amazing Aalto Design Factory where he studied and later taught Product Development and Design Thinking. He now works as a Data Scientist at Sanoma in Helsinki.