Advanced Machine Learning .NET Applications

Introduction

• Advanced machine learning .NET
• What you should know

1. Preparing Data for Machine Learning

• Collecting data correctly
• Utilize SME
• Data types and structure
• Business logic
• Outliers
• Biases
• Data cleansing tools
• Demo: Checking data

2. How to Generate an ONNX Model

• What is ONNX?
• Set up the .NET project for ONNX
• Create a model
• Generate an ONNX model
• Using Netron
• Demo: Generate an ONNX model in Visual Studio

3. How to Utilize TensorFlow Framework

• Image recognition vs. categorization
• What is TensorFlow?
• Set up the .NET project for TensorFlow
• Train the model
• Evaluate the model
• Demo: Train a TensorFlow model in Visual Studio

4. Model Maintenance

• What is MLOps?
• Retraining the model
• Versioning
• Source control

5. Common Pitfalls

• "Machine Learning Model" not in the context menu
• Is 32-bit supported on Windows?
• Ensure the app is targeting x64 or x86
• New project with a different build target
• Challenge: Training and comparing ML.NET models
• Solution: Training and comparing ML.NET models

Conclusion

• Next steps