Deep Learning With C# And CNTK by Mark Farragher

Deep Learning With C# And CNTK

This course will introduce you to Deep Learning and Neural Networks and get you up to speed with Microsoft's Cognitive Toolkit library.

What my students are saying...

‘After hearing about neural networks for years without actually using them, I am proud to say I have successfully trained and used my first neural network – in C#. Thank you so much, Mark. Neural Networks are ridiculously awesome.’
Joel Dokmegang
'I found that participating on such a well-taught course was an awesome experience for me. I think Mark is a gifted teacher as well as amazing technically skilled. It was truly an enlightening experience'
Yoav Kaplan

What you'll learn...

  • Optimizing Deep Neural Networks
  • 1D-Convolution Networks
  • Long-Short Term Memory Networks
  • Style Transfer
  • Generative Adversarial Networks

  • How to train Deep Neural Networks in C#
  • Regression networks
  • Binary and Multiclass Classification Networks
  • Convolutional Neural Networks
  • Recurrent Neural Networks

Course requirements

For this course you'll need:

  • A Windows, OS/X, or Linux computer
  • NET Core version 3
  • Visual Studio Code
  • Optionally: an NVidia GPU
  • At least 1 hour available per weekday

Course description

Are you a CTO, tech leader, or software developer wondering what the Deep Learning hype is all about? Would you like to start experimenting with Neural Networks in your organization?

Then this is the course for you!

In this course, you’re going to master the fundamentals of deep learning in C#. You will learn about building, training, and running deep neural networks with Linear Regression, Binary- and Multiclass Classification. You will build many AI apps with Feed-Forward Neural Networks, Convolutional Neural Networks, and Recurrent Neural Networks. You will also learn how to perform Style Transfer and how to build a Generative Adversarial Network. Your new skills will help you apply Deep Learning to real-world Case Studies.

After completing the course, you will be able to design and train your own neural networks in C# and use them in a large number of problem domains. You will have built price and demand prediction models, medical classifiers for healthcare, a sentiment analysis system, and many other practical AI applications.

What's included?

Video Icon 22 videos Text Icon 81 text files

Here's what's in the course...

Course Introduction
I'm pleased to meet you!
Course prerequisites
3 mins
Installing NET Core 3.0
Installing Visual Studio Code
Introduction To Deep Learning
What is deep learning?
Introducing Linear Regression
Introduction
2 mins
In this section...
Introducing linear regression
10 mins
Linear regression
Introducing multiple regression
13 mins
Multiple linear regression
Introducing regression metrics
12 mins
Regression evaluation metrics
Quiz
Your assignment
4 mins
Assignment: Predict taxi fares in New York
My answers
Recap
Deep Neural Networks
In this section...
Introducing deep neural networks
13 mins
From linear regression to neural networks
The architecture of deep neural networks
How to visualise hidden network layers
How to train deep neural networks
Quiz
Assignment: Predict house prices in California
My answers
Recap
Binary Classification Networks
In this section...
Introducing binary classification
15 mins
Binary classification
Regularization
Introducing binary metrics
14 mins
Binary evaluation metrics
Introducing ROC and AUC
18 mins
ROC, AUC, and Bias
Quiz
Assignment: Predict heart disease
My answers
Recap
Multiclass Classification Networks
In this section...
Introducing multiclass neural networks
7 mins
Multiclass neural networks
Introducing multiclass metrics
15 mins
Multiclass evaluation metrics
Quiz
Assignment: Recognise handwriting
My answers
Recap
How To Train Neural Networks
In this section...
Introducing overfitting and partitioning
12 mins
Overfitting
Partitioning data
Minibatch training
Introducing sparse vector encoding
9 mins
Sparse vector encoding
Introducing K-fold cross validation
4 mins
K-Fold Cross Validation
Quiz
Assignment: Detect spam messages
My answers
Recap
Convolutional Neural Networks
In this section...
Introducing convolution and pooling
21 mins
The convolution layer
The pooling layer
The dropout layer
Introducing data augmentation
8 mins
Data augmentation
Quiz
Assignment: Detect hotdogs
My answers
Recap
Prebuilt Convolutional Neural Networks
In this section...
Introducing feature extraction and finetuning
7 mins
The VGG16 model
Feature extraction
Finetuning
Quiz
Assignment: Detect cats and dogs
My answers
Recap
1D-Convolutional Neural Networks
In this section...
Introducing 1d-convolution and embedding
14 mins
The 1D-convolution layer
The word embedding layer
Quiz
Assignment: Rate movie reviews with an 1D-ConvNet
My answers
Recap
Recurrent Neural Networks
In this section...
Introducing recurrent neural networks
16 mins
Recurrent neural networks
Long short term memory networks
Quiz
Assignment: Rate movie reviews with an LSTM
My answers
Recap
Artistic Style Transfer
In this section...
Introducing style transfer
12 mins
Artistic style transfer
Quiz
Assignment: AI-generated art
My answers
Recap
Generative Adversarial Networks
In this section...
Introducing generative adversarial networks
20 mins
Up-convolutional networks
Generative adversarial networks
Quiz
Assignment: AI-generated fast food
My answers
Recap
In Conclusion
What you've learned