Machine Learning in Ruby

Time & Place: 01.12.2011, 7pm - 9pm at Hacklab 2.0

Meeting agenda: Machine Learning in Ruby

In this presentation Dalibor Nasevic will introduce you to the basic concepts of machine learning. We will look at a prototype implementation of a news aggregator (code can be seen here) that consist of the following components:

training data set
collection data from several news media
parsing of RSS / HTML texts
statistical classification method
creating a cluster with statistical method

We will also look at the different tools used in the implementation of the prototype:

Programming Language: Ruby 1.9.2
Standard libraries: open-uri, net / http, rss (rss/2.0), json
Other Ruby libraries: Sintara, Redis, Haml, Sass, Nokogiri
Database: Redis

Purpose of the presentation:

news aggregator as a concept and realization
brief introduction to the tools used with emphasis on Redis and in what scenarios it can be used
discussion of other algorithms for classification and clusterification of texts (Bayes, Latent Semantic Analysis, K-means
discussion of the architecture for larger applications of this type (couchdb / mongodb, parsing and analyzing texts in Macedonian language)
reviewing ideas for applications of this kind in Macedonia

Истава новост на македонски јазик можете да ја прочитате овде.