We aimed to train our neural network to get as accurate valuation as possible, considering open data available. So, the first task was to mine different types of data from various sources, clean the data and set it up for practical use. During this R&D project we developed proprietary methods for Feature Engineering, thus we can calculate the feature' influence on overall AVM accuracy before implementing all the required work to add it.
ENBISYS performed R&D where the most sophisticated technology of modern time, Deep Learning, has been implemented to its fullest. Unique architecture of neural network, state-of-the-art algorithms, data cleaning methods and feature extracting/engineering have become ENBISYS exclusive competencies that can be translated to other projects in the field of property valuation. This proprietary development will bring the certainty to property purchasers, investors and loan institutions.