Dear community,
i have a set of points with four coordinates (A,B,C,D). One group set are "wrong positives" the other one are "true positives". The question is: how can i find out similarities within each subgroup. Aim is to develop a threshold for a formula (e.g: aA+bB+cC+dD) to predict points group for measured data. Are there regression models?
.
not "regression" as your dependent variable (target variable) is a categorical variable (you have two groups, therefore two categories), rather "classification"
have a look at "discriminant analysis" as a starter !
https://en.wikipedia.org/wiki/Linear_discriminant_analysis
.
Thank you Fabrice,
that is exactly was i was searching for.
Greetings from Munich
Martin
Dear Martin,
The question you ask has been treated in "cluster analysis" and "pattern recognition". You could search in Google.com books or other references with these key words or descriptors to find available sources. Thank you for your question.
https://en.wikipedia.org/wiki/Cluster_analysis
https://en.wikipedia.org/wiki/Pattern_recognition
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics