Context

Prerequisite: Dask Hub platform deployed before.

Dataset: New York Taxi cab statistics simplified.

What has to be done

• What’s wrong with pandas?
• Clean big amounts of data using Dask in the cloud (interactive distributed processing)
• Train a machine learning model with a big dataset in input
• Train machine learning models in parallel (hyper parameter search) on smaller input dataset

Evaluation link

Let’s go

• Just git clone this repository from your Juptyerlab to start
• At the end (end periodically) make sure to save your notebook with outputs by downloading it on your laptop
  • You will be noted based on your notebook content