For all Machine learning & Deep learning people, what was your first classifier model about?

I’m still trying to learn the fucking subject. It’s such a circle jerk of unapproachable material. I swear academics do it deliberately.

My first ever ML thing though was a hand rolled NN in C# a number of years ago to recognise hand drawn digits on a grid. (Use the mouse).

I started with the MNIST handwritten digits classification. That's the 'Hello World' of deep learning!

1. trained an opencv xml based classifier for integration with a sikuli project. played with HAAR, it was too slow, so moved to LBP.
2. then worked on some malware feature extraction and ML project for mobile malware.
Not the classics - but worked for me'

Convolutional neural network that categorised images with cats using tensorflow. Had an accuracy of 97% with regularization. Did it over a UC Berkeley summer camp.
Coming to the field i have realised most fancy models are hard to do in a production environment