I gave a talk at London's Clojure eXchange 2016 on the subject of machine learning in Clojure.
From the programme:
You have heard that Clojure’s ancestor language, Lisp, was developed for artificial intelligence research. Yet, until recently, Clojure’s support for machine learning has mostly consisted of wrapped Java libraries. Informed by his work as a freelance data scientist working primarily with Clojure, Henry will explore some newer libraries that enable sophisticated data analysis unencumbered by Java’s legacy. You will also learn how Clojure’s core sequence-processing strengths naturally lend themselves to simple machine learning techniques you can try at home.