Machine Learning Developer in Cambridge, MA, United States

This project aims to provide basic functionality for solving 2D bin packing problems of irregular (and regular) sets of pieces.

The 2D bin-packing problem consists of a given a set of 2D pieces with unknown form or shape; we have to place them in a series of rectangular bins minimizing the material used. In other words, place all the pieces in as few bins as possible.

The problem is NP-hard, thus there is no guarantee that an algorithm will provide an optimal solution, although we can compare different solutions in terms of how many bins an algorithm has used for a particular set of pieces. This project tries a few heuristics and approximations in order to solve the problem with a reasonable amount of computational effort; given that a brute-force approach to the problem would take virtually an infinite amount of time.

Sample Android code that integrates the Google Maps API for Android with an API. Just a sample code, the API is no longer available.

Code that integrates the Weka ML project with Spark. Written in Java, this is a proof-of-concept code project that seeks to integrate Weka's Machine Learning capabilities with a distributed framework such as Spark.