I have a relatively small dataset (~160) with a mix of data types including continuous, nominal, and binary. I am trying to compute a variable (continuous) from that dataset (multiple data types) that maximizes the likelihood of an outcome X (binary). Ideally, the algorithm would be able to scale to a dataset in the range of the ~1000's.
Chris Hornung Quantitative variables are either discrete or continuous. Variable that is categorical. Categorical variables have a limited number of unique groupings or categories. Categorical data may or may not be in logical order. Gender, material type, and payment method are examples of categorical predictors.
Have a look also https://statisticsbyjim.com/regression/choosing-regression-analysis/
A (mixed-effects) logistic regression is probably a very good starting point:
https://stats.oarc.ucla.edu/r/dae/mixed-effects-logistic-regression/
I am agree with Suddha Sourav's suggestion. For continous or panal data logestic regression is a good choice of model selection, see e.g., https://towardsdatascience.com/building-a-logistic-regression-in-python-step-by-step-becd4d56c9c8
AND
https://www.datacamp.com/community/tutorials/understanding-logistic-regression-python
- Badges
- Science topic