This repo contains code that uses classification to predict build to predict car mileage, given physical attributes of the car. All data used is from UCI Berkeley's public repository. This is a regression exercise that encompasses 'traditional' machine learning and neural networks, with a blog post detailing the effort and results written here: [https://medium.com/towards-artificial-intelligence/machine-learning-and-deep-learning-a-systematic-application-f53b51f3b098]
Defining the problem Data visuaization using seaborn and matplotlib Feature engineering and selection Model fitting and training using ensemble classifiers, decision trees, and the Keras wrapper for neural networks Conclusions and recommendations for client