Summary of Comparative Study on the Performance Of Categorical Variable Encoders in Classification and Regression Tasks, by Wenbin Zhu et al.
Comparative Study on the Performance of Categorical Variable Encoders in Classification and Regression Tasks
by Wenbin Zhu, Runwen Qiu, Ying Fu
First submitted to arxiv on: 18 Jan 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: None
GrooveSquid.com Paper Summaries
GrooveSquid.com’s goal is to make artificial intelligence research accessible by summarizing AI papers in simpler terms. Each summary below covers the same AI paper, written at different levels of difficulty. The medium difficulty and low difficulty versions are original summaries written by GrooveSquid.com, while the high difficulty version is the paper’s original abstract. Feel free to learn from the version that suits you best!
| Summary difficulty | Written by | Summary |
|---|---|---|
| High | Paper authors | High Difficulty Summary Read the original abstract here |
| Medium | GrooveSquid.com (original content) | Medium Difficulty Summary Machine learning practitioners often struggle to choose the right encoder for categorical variables in their datasets. This is because different encoders can have a significant impact on model performance. In this study, researchers categorize machine learning models into three types: those that perform affine transformations (ATI), tree-based models, and others. They theoretically prove that the one-hot encoder is the best choice for ATI models, while the target encoder is most suitable for tree-based models. The study also conducts comprehensive experiments to evaluate 14 encoders on 28 datasets with eight machine-learning models. The results agree with the theoretical analysis, providing valuable insights for data scientists in fields like fraud detection and disease diagnosis. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper helps us choose the right way to turn words into numbers for our machine learning model. It shows that different methods can work better or worse depending on what kind of model we’re using. The researchers tested 14 different ways to do this on lots of datasets and showed which ones work best with certain types of models. This can help us make better choices when working with data. |
Keywords
* Artificial intelligence * Encoder * Machine learning * One hot
