Clustering is done based on houses data and say suppose cluster 1 contains 5 houses. Appliance 1 is used in 2 houses out of 5 and appliance 2 is used in 4 out of 5 houses.Each appliance produces an error . Need to find sum of the errors of each appliance in each cluster. Summing up of errors at appliance level would not give us the correct picture as appliance 1 is used in 2houses but appliance 2 is used in 4houses. How do we handle such data? How we get the total errors produced from each appliance for a particular cluster?

  • $\begingroup$ one way would be to count errors of appliance only where it is used (ie only on 2 houses) and so on.. $\endgroup$ – Nikos M. Jan 17 at 8:15
  • $\begingroup$ Doing this, we would count errors of one applaince for 4 houses n the other for 2 houses. For example, an appliance used in 2 houses might give more errors when compared to the appliance used in 4houses. So shouldn't we calculate the errors for all the houses as there are 5 houses in the cluster ? If yes, I would like to know if there is a way to calculate errors for the appliance which is in 2houses ..what would be the errors if it was used in other three houses in the cluster as cluster has 5houses in total.. $\endgroup$ – User06 Jan 17 at 11:46
  • $\begingroup$ Similarly, for the all the houses which has less than the size of the cluster $\endgroup$ – User06 Jan 17 at 11:48
  • $\begingroup$ You make it very convoluted, other approach is use a weight (eg $\frac{n_{houses}}{n_{cluster}}$) $\endgroup$ – Nikos M. Jan 17 at 16:38

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.