text mining new set of data in production environment expect training feature

Question

Hi,

I have trained a model based on 28000 features but now when I am introducing the new set of data for prediction it's expecting the same feature.

Could someone help me on this, please?

Thanks

Comments

Hey @manu, can you please share some more details?

---

Hi Karthik,

I am working for text classification multi-class classification and in my training data on which my model is trained is having a feature of 28000. My question is when I introduce a new data set for prediction why predict function expect the same list of the feature on which model trains.

I a using R and model is done using decision tree .

---

Hi @Many, if you have built the model based on these 28000 features, the model will only be able to predict data for what it's trained for. Does the new data set that you are introducing have a correlation with the training dataset?

Also, share more details based on the datasets that you are using.