Implementing a new decision support system for housing market analysis using adaptive and attention-based residual RNN with structural equation modelling

A housing property is a shelter that investors purchase for investment purposes or to protect themselves from potential threats. Deep learning-aided housing market prediction models are beneficial in real estate trade market. A new decision-making system for housing market analysis is developed in this research work using a deep learning technique. In initial phase, the required data for validation are manually collected through questionnaire responses, which are processed using Structural Equation Modelling (SEM). Next, the collected data is offered to housing market decision making phase. In this phase, Adaptive and Attention-based Residual Recurrent Neural Network (AA-RRNN) is employed to carry out housing market decision-making procedures. The parameters in AA-RRNN are tuned using Leader-Driven Quokka Swarm Optimization (LD-QSO), which helps improve the decision-making efficiency. Finally, housing market decision-making outcomes are obtained from AA-RRNN. Later, various experimental validations are carried out in the developed housing market decision-making system by considering various conditions.