Machine Learning Guide

Steps

ML Workflow

Data Exploration

Data Properties and Data Visualization

  • data analysis to get a feeling for the dataset
  • check the missing data
  • learn which features are important

Data Preprocessing

  • compute missing values
  • convert features into numeric ones
  • group values into categories
  • create a few new features

Train a Model

  • train somen model
  • pick the best one and apply cross validation on it
  • opimize hyperparameter values
  • assessment
posted @ 2018-07-29 11:55  BerMaker  阅读(123)  评论(0)  编辑  收藏  举报