High Performance deep learning with Intel technologies

Shailen Sobhee

Cross-Platform-Development Data Science Deep Learning Distributed Systems Performance

See in schedule Download Slides

We take a look at a solid use-case of the application of deep learning in the healthcare sector. In a hands-on fashion, you will get the chance to walk through the Python-based algorithm that performs a tumour segmentation from MRI scans of the human brain. We will explore the deep neural network used, the framework leveraged and the hyper-parameter tuning considerations for best training and inferencing performance. In this session, you will also see what the Intel® Math Kernel Library for Deep Neural Networks (Intel® MKL-DNN) does behind the scenes to maximise hardware performance.

Note: Basic understanding of neural networks would be ideal. Knowledge about Tensorflow+Keras is a bonus.

Type: Training (180 mins); Python level: Intermediate; Domain level: Intermediate

Shailen Sobhee


Shailen is an AI specialist at Intel. He is the link between the core software engineering team and Intel's end-customers. In his role, Shailen assists and trains customers on adopting the latest and greatest optimized machine-learning and deep-learning frameworks in their software development process. Shailen holds a Master’s degree in Computational Science and Engineering from the Technical University of Munich.

Shailen plays the piano and is an avid football player.