Summary of Naive Bayes Classifiers and One-hot Encoding Of Categorical Variables, by Christopher K. I. Williams
Naive Bayes Classifiers and One-hot Encoding of Categorical Variables
by Christopher K. I. Williams
First submitted to arxiv on: 28 Apr 2024
Categories
- Main: Machine Learning (cs.LG)
- Secondary: Machine Learning (stat.ML)
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 The paper explores the consequences of incorrectly encoding categorical variables as one-hot vectors when using Naïve Bayes classifiers. This mistake leads to a product-of-Bernoullis assumption instead of the correct categorical Naïve Bayes classifier. The differences between these two approaches are analyzed mathematically and experimentally, revealing that they often agree on maximum a posteriori class labels, but with varying posterior probabilities. |
| Low | GrooveSquid.com (original content) | Low Difficulty Summary This paper looks at what happens when we wrongly represent categories as numbers using one-hot encoding in Naïve Bayes models. This mistake leads to different results than if we used the correct categorical approach. The researchers compare these two methods and find that they usually agree on the most likely outcome, but with varying levels of confidence. |
Keywords
» Artificial intelligence » One hot
