I have a feature called gender in my model and it's has 3 values male, female, child. When I trained the model I converted it to one hot encoded and then trained the random forest classifier and saved it. Now when I load the classifier and try to predict for a single inference I provide 1 value for ex. Male and one hot encode it..the classifier says the input feature is a mismatch, missing values for features (child, female). How to resolve this issue...This is just an example in original feature there are more values for categorical columns.. could someone please help.
PS I am not using label encoder because I don't have any ordering for the column and I don't want to use it.