PinnedInTDS ArchivebyHuey Fern Tay·Jul 25, 2021Using R’s simputation to handle missing valuesHoles in our dataset can be solved by making intelligent inferences from a linear modelA response icon1A response icon1
IncodinginrbyHuey Fern Tay·Jan 12, 2022Detecting multicollinearity — it’s not that easy sometimesBy Huey Fern Tay
InCodeXbyHuey Fern Tay·Jan 6, 2022How to analyze the NAs in multiple response surveysBy Huey Fern Tay, With Greg Page
Huey Fern Tay·Nov 18, 20215 things I learned by analysing Kaggle’s 2020 Machine Learning and Data Science SurveyBy Huey Fern Tay
InTDS ArchivebyHuey Fern Tay·Oct 4, 2021A Little-Known Trick in Hierarchical Clustering: WeightsRecalibrate your dataset to answer a business questionA response icon2A response icon2
InTDS ArchivebyHuey Fern Tay·Aug 11, 2021When is it ok to impute missing values with a zero?Substituting missing data with a zero can sometimes be justifiedA response icon1A response icon1
InTDS ArchivebyHuey Fern Tay·Jul 9, 2021Substituting missing data with the group average — why it’s good to be cautiousBy Huey Fern Tay
InTDS ArchivebyHuey Fern Tay·Jul 3, 2021When a crazy wind wipes away our records — what do you do?By Huey Fern Tay