Optimizing Classification problems for Parallela

implement a parallel optimized algorithm for a classification engine related to machine learning and demonstrate a solution to an interesting classification problem

parallella cluster

Parallella Project – Optimizing Classification problems for Parallella

The project will include two main stages –

(I) The first stage aim is to familiarize with the `parallella` system and the methods for programming its multi-core capabilities. In this stage you will be required to implement a parallel sorting algorithm using
at least two methods of parallel programming. Among the preferred technique for programming which you can use are:
– OpenCL
– MPI type of message passing
– OpenMp

    – Customized solution  or others.
(II) In the second stage you are required to implement a parallel optimized algorithm for a
classification engine related to machine learning and demonstrate a solution to an interesting classification
    problem,
The classification implementation should be optimized for maximal performance and you will be required to evaluate
the performance by benchmarking and comparisons to several types of implementations.