TOPICS COVERED IN THIS LECTURE
This lecture covers Deep Neural Networks, data Scaling, One-Hot Encoding, and using Training And Validation Sets to train a neural network. My C# code uses Accord to build and train a neural network, and Deedle for data processing.
The great thing about machine learning is that there are always a ton of challenges out there. Governments, companies, and research labs are constantly introducing cool and complex new problems for computers to solve.
As a machine learning developer, you will never be bored 😉
Here is a classic machine learning challenge I’d like to share with you. It’s called ‘MNIST’ and uses the following training data:
Your goal is to build an app that can read these handwritten digits, and correctly associate each one with the digit value it represents.
So basically, what you’re doing here is Optical Character Recognition (OCR) on handwriting. Something that software developers have struggled with for decades.
But did you know you can easily solve the challenge in C#?
Watch the lecture and I’ll show you how to build a deep neural network in C# that can recognize these handwritten digits and correctly transcribe them to text.
– What is machine learning?
– Popular C# computer vision libraries
– The challenge: recognizing handwritten digits
– Introduction to neural networks
– The neural network we’re going to use
– Training, validating, and testing neural networks
– C# code walkthrough
– Questions & answers
‘I’m freshly new on Neural Networks and Machine Learning. I saw this video which is a nice way to shine the light on Neural Networks in C#. Great work!