The 'Deep learning fundamentals' track

What is the Deep learning fundamentals learning track?

The Deep learning fundamentals track focuses on solving common computer vision and natural language processing tasks with deep neural network architectures using PyTorch.

What trainees will learn:

Trainees will solve several CV (computer vision) and NLP (natural language processing) tasks during the training. The curriculum of this learning track is heavily practice-driven. The syllabus is customizable to the business requirements and background of trainees; the tasks are set by us or mutually defined according to your requirements.

What trainees will take away, i.e. our deliverables:

• Python code templates for solving the tasks specified in the curriculum
• Lecture notes
• DL textbooks
• Optional: TeamHub, the codebase & documentation knowledgebase

Pre-requisites

• Basic Python programming proficiency
• Prior experience with machine learning
• Linear algebra fundamentals

Typical background: mid-level and senior data scientists with solid background in mathematics and programming who seek to extend their expertise and solve CV and NLP tasks. Development of predictive models is part of the learning track. Specific business tasks to which the training shall be tailored are to be defined.

Light-weight versions of the Deep learning fundamentals track can be customized to match data scientists’ skillsets. The tasks to be solved as part of practice-based projects may be not as advanced but sufficient to introduce the trainees to real-life applications of deep neural network architectures.

Why Deep learning fundamentals learning track?

Deep Learning is transforming multiple industries. This deep learning track will help learn DL fundamentals, understand and use modern architectures and build systems in CV and NLP.

The Deep learning fundamentals syllabus

* Intro: Torch and deep learning
* Convolutional Neural Networks
* Recurrent Neural Networks
* Sequence2Sequence
* GAN, Generative Adversarial Networks
* Variational Auto Encoder
* DeepSpeech
* Gaussian Mixture Models
* Reinforcement Learning
* Neural Processes

* Transformer
* Convolutional Seq2Seq
* Memory Networks
* Neural Turing Machine
* Question Answering
* Abstractive Summarization
* Neural Dialogue
* Unsupervised Machine Translation

* Mask R-CNN
* U-net
* YOLO
* Mobile Nets
* Stack GAN
* Cycle GAN
* Pixel CNN/RNN
* Visual Dialogue
* Unsupervised Computer Vision

* Adversarial Network based Recommendation
* Representation Learning from User and Item Content Information
* Deep Explainable Recommendation