Given a dataset with 30 fields and 25000 instances,

1) what are your suggestions for novel methods of rule extraction?

2) Can I use association rule mining in addition to sequential rule mining?

3) which method can be more appropriate for such a big dataset?Apriori-based like SPIRIT, SPADE, SPAM, IBM or pattern growth ones like FreeSpan, PrefixSpan, SLPMiner?

*the output field (risk) is labelled as very low, low, medium, high, very high. Also, there is a temporal field (date) for each instance. An example is given below.

$$ \begin{array}{|c|c|c|c|} \hline \mathbf{date}& \mathbf{temperature} & \mathbf{density} & \mathbf{risk} \\ \hline 2018/1/2 & 15 & 100 & \text{very high}\\ \hline & & &\\ \end{array} $$

  • $\begingroup$ Hi Anna, What do you mean by novel? There are methods to extract rules, yes. Without any given target variable you could use both methods to find correlations in the dataset. $\endgroup$ Apr 1 '19 at 22:15
  • $\begingroup$ Hi Balen, by the novel, I meant, for example, using deep neural networks or those that can work with my large dataset. $\endgroup$
    – anna
    Apr 2 '19 at 17:08

Your Answer

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

Browse other questions tagged or ask your own question.