Method for Providing Real Estate Estimated Real Transaction Price Calculation Service Using Decision Tree Based Time Series Trend Prediction Learning Model

摘要

The method of providing a real estate expected real transaction price calculation service according to the present invention comprises the steps of: establishing a database of actual transaction prices; separating data before a specific time into a training set and data after a specific time into a test set; generating an instance, and gradient boosting to generate a decision tree-based time series trend prediction learning model; verifying the accuracy of the learning model by measuring an error between predicted virtual data and actual transaction price data by applying the feature instance of the test set to the learning model; Further expanding the learning tree when the target performance is not achieved according to the accuracy verification result, and fixing the learning model when the target performance is satisfied; and providing real estate actual transaction price prediction information obtained by applying feature instance data corresponding to the request of the user terminal to the fixed prediction model to the user terminal.