ML Project Checklist - liniribeiro/machine_learning GitHub Wiki

• Analize the problem;
• Get data;
• Identify and visualize the data to get usefull infos;
• Prepare the data to be used by the ML algorithm;
• Select and train a model;
• Improve model;
• Present the solution;
• Make available in production, monitoring and maintain.

Analyzing the problem

• Whats the business objective?
• Whats the actual solution?
• We are going to use supervised, not supervised or reinforcement learning?
  • Every time that we have labeled data, we can use the supervised learning
• its going to be a task of regression or classification? or another?
  • Every time that you are asked for predict a value, its a regression task.
• We are going to use training in batches or online?
  • If we don't have a large amount of data and we don't have the necessity of using newer data, we use the training in batches.
  • If the data are very bug, we can use the MapReduce technique or use online training.