1
$\begingroup$

I am looking for algorithms to find pattern or more precise correlations in lists compared to an Output. Let us assume I have a Database like this:

  1. Input: [A,C,D,E...], Output: Positive
  2. Input: [A,B,C,E,F...], Output: Negative

The Problem is that the distinct Input values are roughly 1000 and not just 6 like in my example (A-F). The output is binary though.

Do you know of any algorithm that detects correlations in the Inputs to finally detect the most critical Inputs that lead to a Positive Output?

$\endgroup$
1
$\begingroup$

You could start off with something as simple as logistic regression to model your problem. You could then experiment with Random Forest Classifiers and then graduate to CNNs and RNNs.

$\endgroup$
1
  • $\begingroup$ Could you elaborate on that please? How am I supposed to define the chosen Feautures? I would have like 500 features.. how to handle them? $\endgroup$
    – can.inan
    Jun 5 at 14:55
1
$\begingroup$

It clearly states that you are dealing with simple classification problem. So you don't need to go with CNN you can use Machine learning classification Algorithms like

  1. Logistic Regression.
  2. SVM
  3. KNN (k-Nearest Neighbor)
$\endgroup$
1
  • $\begingroup$ No It is NOT a simple ML problem.. it is very tough because I Have Lists as input.. If I transfer them to columns as Inputs it will be a lot plus they will be binary (1 or 0).. and then i will need to do some feature selection $\endgroup$
    – can.inan
    Jun 6 at 20:42

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.