Building ML Web Applications

A guide to getting your ML model to work as part of a web application.

Transcript

2. Python Powered Disease Classification Web Applications
3. Daniel Madalitso Phiri organiser @devcon_zm malgamves
4. I ❤ engaging crowds Note:
5. Python Powered Disease Classification Web Applications
6. So Machine Learning? Making computers do things without explicitly programming them to
7. My Story
9. You probably get it, Machine Learning is cool..
10. It’s solving some really complex problems.. It’s ok
11. Let’s talk about notebooks Jupyter Notebooks
12. Let’s talk about lungs Bad lungs to be specific
13. We have a massive labelled data set and an accuracy problem What do we do?
15. Notebooks are great, lots ask. “How would this work in the real world?”
16. I'm here to show you how. TL;DR - This talk is about Flask and Tensorflow and OpenCV
17. How will we ever train a successful model without the compute power?
18. Mind Map - let’s think about what a solution looks like Zen:
19. - App architecture - Implementation - Deployment Content:
20. - Upload form - Space to display results Client:
21. Demo
22. So you want to be able to deploy your app and have a CI pipeline. Server:
23. - Updating the model needs to be easier Problems:
24. This is my implementation, yours could be different. Identify your domain
25. bit.ly/easymlapp Let’s Embrace OSS
26. Thank You bit.ly/feedmalgamves