why is this a mistake?

See more:

SQL-Server-2005

SQL-Server

BI

I am doing a data mining job, but there are many mistakes I don't know how to solve them.
Firstly, I created a mining model: decision tree:

create mining model [DecisionTree_MemberLevelPredict]
(
	stu_ID text key,
	stu_gender text discrete,
	stu_birthyear text discrete,
	stu_education text discrete,
	stu_levels text discrete predict
)
using microsoft_decision_trees (complexity_penalty=0.5)

and then I try to train the model:

Insert into [DecisionTree_MemberLevelPredict]
(stu_ID,stu_gender,stu_birthyear,stu_education)
OPENQUERY([Bookstore],
   'SELECT ID,gender,birthyear,education,levels
    FROM dbo.customer')

and at last I want to use the model to predict the new people in table customer,
and I face the mistakes below:
The predict DMX code

select T1.ID, DecisionTree_MemberLevelPredict.stu_levels, PredictProbability(DecisionTree_MemberLevelPredict) as Proba
from DecisionTree_MemberLevelPredict
prediction join 
OPENQUERY([Bookstore],
   'SELECT ID,gender,birthyear,education,levels
    FROM dbo.customer')as T1
on DecisionTree_MemberLevelPredict.stu_ID=T1.ID and DecisionTree_MemberLevelPredict.stu_gender=T1.gender and DecisionTree_MemberLevelPredict.stu_birthyear=T1.birthyear and DecisionTree_MemberLevelPredict.stu_education=T1.education

it says "T1.ID grammar" is not right, but ID is one column in my table customer and T1, how is that happen? What I did wrong?

Thank you very much !

BTW, after I have created the mining model, it automatically generate one mining structure, but I could not find the mining structure in my bi dev.