First, all the classification model will work for this real state classification. Only the thing matters is accuracy. there are several classification algorithms that can implemented such as decision tree, random forest, k nearest neighbor, Naive Bayes, svc etc. Among all of them, mine favorite is random forest.
I have implemented the decision tree classifier for the given data, and it is giving me an accuracy of 93% which is very good. Another very important thing is handling the data. Which data is important, and which is not is also very important. The categorical data handling, the ordinal and nominal, handling the nans and various other things matters in the process of feeding the model the correct data so that the output is not hampered.
If there are few more data than may be the accuracy can be improved and thus data plays a vital role in accuracy.
The conclusion is all the model will work fine if data feeding to the model is good. Thus, a keen knowledge of data is used.