Train a car condition classifier from scratch

This is a step by step tutorial of training a neural network classifier. The dataset I'm using is from UCI. I've also wrote a detailed tutorial in Chinese in 莫烦Python.

Data description

4 Classes about car's condition:

• unacc: unaccepted condition
• acc: accepted condition
• good: good condition
• vgood: very good condition

Features:

• buying: vhigh, high, med, low.
• maint: vhigh, high, med, low.
• doors: 2, 3, 4, 5more.
• persons: 2, 4, more.
• lug_boot: small, med, big.
• safety: low, med, high.

Training

Files:

• data_processing.py : download data and process to an accepted format
• model.py : training model and view result

Dependencies

• Python
• tensorflow
• pandas
• numpy
• matplotlib

You can view more tutorials on this page or know more about me on here.