Advanced data science systems and methods useful for auction pricing optimization over network
An advanced data platform may receive an asset pricing request containing information about an asset. An optimization engine may determine a predicted price for the asset at different locations and times and compute a price matrix accordingly. The engine may identify an optimized predicted price from the price matrix, taking into account the spatial and temporal factors and various optimization conditions. A view for presentation of the optimized predicted price for the asset on a client device is generated and communicated to the client device over a network. When the asset is a vehicle, the engine may compute a linear regression model that defines a set of input variables with associated regression coefficients, the set of input variables comprising input variables representing attributes describing the vehicle.
1. A method for generating a quantitatively optimized auction price prediction, the method comprising:
receiving, by a computer from a user device, an asset pricing request containing information about an asset;
computing, by the computer, a price matrix for the asset, the price matrix containing a plurality of values for a predicted price of the asset, each value of the plurality of values associated with a specific location of a plurality of locations at a specific time of a plurality of times;
iteratively analyzing elements of the price matrix, the iteratively analyzing performed by the computer with respect to the plurality of locations and the plurality of times, the iteratively analyzing the elements of the price matrix including:
comparing a predicted auction price associated with a first time of the plurality of times to a predicted auction price associated with a second time of the plurality of times for selling the asset at a given location;
responsive to the predicted auction price associated with the second time of the plurality of times for selling the asset at the given location being larger than the predicted auction price associated with the first time of the plurality of times for selling the asset at the given location, determining that a first optimization condition is met;
identifying an optimal location-keyed auction price, given that the first optimization condition is met;
comparing a predicted auction price associated with a first location to a predicted auction price associated with a second location for selling the asset at a given time;
responsive to the predicted auction price associated with the second location for selling the asset at the given time being larger than the predicted auction price associated with the first location, determining that a second optimization condition is met;
identifying an optimal time-keyed auction price, given that the second optimization condition is met;
comparing a predicted auction price associated with a first time and a first location to a predicted auction price associated with a second time and a second location for selling the asset;
responsive to the predicted auction price associated with the second time and the second location for selling the asset being larger than the predicted auction price associated with the first time and the first location, determining that a third optimization condition is met;
identifying an optimal time-and-location-keyed auction price, given that the third optimization condition is met; and
in response to the asset pricing request, returning, by the computer to the user device, at least one of the optimal location-keyed auction price, the optimal time-keyed auction price, the optimal time-and-location-keyed auction price, or a baseline price as an optimal auction price of the asset.