Intelligent prediction method based on big data

Abstract

The invention discloses an intelligent prediction method based on big data. The intelligent prediction method comprises the following steps of step 1, data cleaning; preprocessing the original data; step 2, characteristic engineering; analyzing the cleaned data, finding out important features, and constructing statistical features related to the important features; step 3, constructing a model; anXGboost model, a LightGBM model and a catboost model which are high in classification precision are used for predicting the rent; step 4, model fusion; according to the score of the model predictionresult, different weights are given to the model prediction result, a final model is formed through weighting, and the model generalization ability is improved; and step 5, intelligent prediction; accurate prediction of the rent is realized by utilizing the obtained model; according to the invention, by use of data such as real estate markets, rental markets, market demands and house configuration, by use of model algorithms such as machine learning and artificial intelligence, and in combination with the innovation capability of the model, reasonable house rental prediction is made so as to cope with the influence of market changes on operators and real estate institutions.

Description

technical field [0001] The invention relates to the technical field of big data and machine learning, in particular to an intelligent prediction method based on big data. Background technique [0002] In recent years, the domestic housing rental market has entered a new stage of development, and the long-term apartment market, as an important part of the rental market, has received more and more attention. But at the same time, China's long-term apartment market is also facing multiple challenges such as enterprise market entry, business (store) expansion, capital market games, corporate restructuring and mergers and acquisitions, etc. Among them, how to accurately predict rents has become a major problem in the development process of the industry; The determination of the rent needs to be determined in combination with various factors such as the real estate market, the leasing market, market demand, location and location, and housing configuration. However, there is no mea...

AI Technical Summary

Problems solved by technology

But at the same time, China's long-term apartment market is also facing multiple challenges such as enterprise market entry, business (store) expansion, capital market games, corporate restructuring and mergers and acquisitions, etc. Among them, how to accurately predict rents has become a major problem in the development process of the industry; The determination of the rent needs to be determined in combination with various factors such as the real estate market, the leasing market, market demand, location and location, and housing configuration. However, there is no means of predicting the rent of relevant houses in the prior art, and the renter cannot effectively predict market changes. , so as to make a reasonable decision

Images