Boston house prediction

#1

i am confused how can i process the given data to feed it to NN ?
becaue in case of IMDB sentiment we convert the strings into vectors but in this case i dont think we have to do that so what should i do?

#2

Hey @footballer.salik, in boston house prediction data all features are numerical. So no need of such type of pre processing, but you may need to perform normalization.

In Strings in imdb sentiment there were words, so we need to find a way out to convert them into numerical form and that’s the reason we converted them into vectors. So here’s no need to do any such things since all features are already in numerical form.

Hope this cleared your doubt :blush:
Happy Learning :sunny:

#3

but features like housingStyle ,neighbourhood,street and many more have strings value what about them?
i cant exclude them because features like neighbourhood is important for the price.

#4

Yes one way is to simply exclude them, but in case if you want to include them as well, than you need to your sklearn Label Encoder or One hot vector. Both of these functions are in sklearn library, you can use them directly from there.

Its totally upto you, since this is part of feature engineering.

Hope this cleared your doubt. :blush:

#5

How accurate are those house predictions? And is it more related to sociology or the economy? One of my friends needs to do a forecast like this as his home assignment, and he doesn’t know what to start with and what data to use.

#6

Can it be helpful while choosing a house in an area? I am looking for a house to buy, and I would appreciate any good sources of information about which town
or neighborhood to choose and where are the best prices. The fact is that I plan to buy a house with the help of Equity Release Durham and then sell it in around 10 years to get some profit and get a bigger house. So I need to buy a house in an area where the housing price is expected to rise.