Join the Training

Do not miss out on the most updated Deep Learning: Recent Advances and Applications Training!

About the Course:

Deep Learning applies multi-layered neural network architectures for solving complex problems in computer vision, natural language processing, genomics, gaming, robotics, and many other areas. Two actively researched architectures in deep learning at the moment are Convolutional Neural Networks and Deep Recurrent Neural Networks. Other topics of interest in Deep Learning are Deep Belief Nets, Deep Reinforcement Learning and Generative Adversarial Networks. This workshop will offer a gentle introduction to Deep Learning, with the just the right level of algorithmic and mathematical detail, and more focus on programming tools and demonstration of applications. The goal is to enable Developers, Technical Managers, and Business Leaders, translate the buzz and jargons in this field to a clearer understanding of how these technologies work, and what technical and business problems they help us solve. The workshop is designed in such a way that the attendees can gain a fairly wide-ranging overview of Deep Learning, and practical insights, within a day. In this workshop we will use Tensorflow as the Deep Learning platform, and the entire set-up including source code for the applications will be delivered in a Virtual Machine.

Pre-requisites:

A basic awareness of machine learning and neural networks will suffice. No prior technical knowledge in this field will be assumed. Some programming experience (in any language) will be assumed.

Participants should bring their own laptops - they will get a chance to run some of the demos, with trainer's guidance.


Course Structure

Course Objective:

Neural Networks: The Journey from Feed-forward to Convolutional

  • Overview of Feed-forward Neural Networks
  • ImageNet and the Emergence of Convolutional Neural Networks
  • CNN Architectural Variations (including architectures deployed at Microsoft and Google)
  • Application: Captioning a Large corpus of images using CNN (with demo and code walkthrough)

Introduction to Deep Belief Networks

  • Belief Networks and their application to Causal Analysis
  • Application: Medical Diagnosis using Deep Belief Networks

Recurrent and Deep Recurrent Neural Networks

  • Recurrent Neural Networks: Algorithm and Processing of Sequential Information
  • Deep Recurrent Neural Network Architecture
  • Application: Deep Recurrent Neural Networks for Natural Language Processing

Introduction to Deep Reinforcement Learning

  • What is Reinforcement Learning?
  • Using Deep Learning to solve Reinforcement Learning Problems
  • Application: Deep Reinforcement Learning in Gaming Applications

Register

DISCLAIMER AND INDEMNITY

Confirm your CANCELLATION in writing up to 15 working days before the event and receive a refund less a 10% service charge. Regrettably, no refunds can be made for cancellations received less than 15 working days prior to the event.

However, SUBSTITUTIONS are welcome at any time and is done at no extra cost. The organisers reserve the right to amend the programme if necessary.

Important Disclaimer:The organizers reserve the right to make substitutions or alterations and/or cancel a speaker(s) if deemed necessary by circumstances beyond its control.

INDEMNITY: Should for any reason outside the control of UNICOM Training & Seminars (P) ltd (hereafter called UNICOM), the venue or the speakers change, or the event be cancelled due to industrial action, adverse weather conditions, or an act of terrorism, UNICOM will endeavour to reschedule, but the client hereby indemnifies and holds UNICOM harmless from and against any and all costs, damages and expenses, including attorneys fees, which are incurred by the client. The construction validity and performance of this Agreement shall be governed by all aspects by the laws of India to the exclusive jurisdiction of whose court the Parties hereby agree to submit.