0
$\begingroup$

have a question about the type of model which I should use for a dataset I have.

I have use 2 data-sets for my project. After hypothesis testing, I me

Out of the 7 input variables, 6 of them are categorical and 1 is a date column. Now I have encoded the categorical columns using label encoding and converted them into numerical values. Now I’ve used a simple linear regression model on this dataset and achieved a normalized RMSE value of 0.11.

If I want to improve this accuracy, how do I go about doing it? How can I derive upon the kind of models that I can use considering the data set I have?

The data is mostly about forecast revenues for each product group. That's why I have those categorical columns which represent the hierarchy of the product groups.

$\endgroup$
  • $\begingroup$ If it is a time series with not too many points in time, you may encode time as dummies (indicators), e.g. as week or day of year etc. You can use this in a linear regression, possibly with Lasso (to shrink irrelevant features). If it is a dense time series with sufficient observations, you could try a LSTM neural net. $\endgroup$ – Peter May 29 at 14:44
  • $\begingroup$ I am new user on stack Exchange , so i dont know how to edit question. My original question is: I have two datasets for my research. AFter hypothesis testing, I have merged two datasets into one dataset. In dataset, I have three columns: first column representing years; second column described crime rate and third column represented unemployment rate . My research question is "How unemployment rate affect crime rate in Ireland?" Could you please suggest me which model I shroud use for prediction? $\endgroup$ – Ankit Patel May 29 at 14:44
  • $\begingroup$ @Peter please read my original question in comment and give me your valuable input. $\endgroup$ – Ankit Patel May 29 at 14:46
  • $\begingroup$ The research question sounds more like a case of causal modelling. You should look for drivers of crime, esrimate over years, and have a look what the marginal effect of unemployment is. The reason is that many other things too explain crime. So I guess that you will not be able to make great predictions based on the data you have. From the structure of the data, however, I think OLS (linear regression) is a reasonable method. $\endgroup$ – Peter May 29 at 15:12
  • $\begingroup$ @Peter Thank you so much for suggestion. $\endgroup$ – Ankit Patel May 29 at 19:16

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.