ROOT provides via the TMVA package a new Deep Learning framework, which allows to build complex Machine Learning architectures which can be used for data classification or regressions. In addition to standard dense layers, this new Deep Learning framework supports also convolutional or recurrent layers. The candidate will be working for TMVA, with the main goal to optimize the performance to evaluate trained deep learning methods and to compare the performance results with other existing Machine Learning libraries. Optimal performances are crucial for real time exploitation of Machine Learning tools in High Energy Physics.
C++ skills, experience with large scale software development, vectorization, linear algebra libraries and profiling tools is a plus.