Hello everyone, I'm seeking some advice or references related to the optimal number of observations needed per category within a categorical variable for machine learning projects. I've come across a rule of thumb suggesting that a minimum of 20 observations per category is advisable. However, I'm curious about the community's views on this and whether there's any literature or research that could provide more detailed guidance or confirm this rule. Any insights or recommendations for readings on this topic would be greatly appreciated. Thank you!
Dear Guillermo Palchik ,
Generally speaking, the rule of thumb regarding machine learning is that you need at least ten times as many rows (data points) as there are features (columns) in your dataset. This means that if your dataset has 10 columns (i.e., features), you should have at least 100 rows for optimal results.
https://graphite-note.com/how-much-data-is-needed-for-machine-learning#:~:text=Generally%20speaking%2C%20the%20rule%20of,100%20rows%20for%20optimal%20results.
Regards,
Shafagat
Thank you Shafagat Mahmudova,
My question was about something a bit different, though. If I'm dealing with a column representing a categorical variable—for instance, "color", which includes three options: red, green, and blue—how many entries or rows would be necessary for each of these categories: red, green, and blue?
- Badges
- Science topic
- Similar topics
- Filing