Today there are many offers for real estate investments. Today, for those who invest in real estate for the purpose of reselling this property, you have to use the databases of various auction companies, banks, for example, such as Zillow, Foreclosure.com, Auction.com and others.
At the moment, our company is developing its database of real estate objects, which will allow you to find possible objects requiring repair or reconstruction for the purpose of further resale.
We use modern tools and platforms to create an algorithm that allows you to determine real estate, with an untidy territory or problems in the construction of the roof, etc. The market value of such real estate is much less, since not always this object can be put up for standard sale. Having bought the property at a low cost, having made repairs, the investor can sell this property and make money on it.
Our algorithm is based on a convolutional neural network that analyzes images from space provided by landsat_8. The algorithm is trained to find houses with an untidy territory, with problems of the roof and other structures.
After analyzing the data, a database of addresses of real estate objects is formed.
Perhaps this database of real estate will be interesting not only to investors, but also to construction companies.