It is nowadays seen that accurate cost estimation at the early stage plays an important role in any initial construction, as it affects the success of the project and that underestimating and exaggerating the cost may lead to the failure of the project. In terms of costs. That is, it may lead to a loss or shutdown of the project due to unavailability of cost despite the limited design data available at that stage. Any project's completion is defined by better quantitative and cost estimation techniques that allow for the most efficient utilization of resources. The goal of this project is to develop a cost estimation technique that uses an artificial neural network (ANN) model developed by Long Short Term Memory (LSTM) to predict the total structural cost of buildings by considering various parameters and to quickly and accurately predict the estimated cost at an early stage. Estimators employ a variety of techniques to forecast the cost of construction and civil engineering projects. The goal of this research is to use Artificial Neural Networks to come up with a cost estimate (ANN). The accuracy of these models, like those created by LSTM, is ultimately determined by a realistic estimated value. Microsoft Excel Solver was used to construct a neural network model analysis. As a result, the goal of this study was to design a model for estimating the cost of construction projects in the early stages, where data was collected for 120 distinct structures and a number of relevant cost criteria were discovered. From the buildings, by distributing a questionnaire to a number of engineers for consultants in this field, the neural network was trained and tested through a set of designs. The proposed architecture was adopted, and the optimum neural network architecture with minimal error. The income layer of the proposed neural network consists of four variables, and the results obtained from the neural networks indicated the success of this technique in predicting the cost of construction projects in the first phase of the project, based on a set of information without the need for more details related to the design. Simple program development using an easy-to-train neural network. That is, the neural networks were reasonably successful in estimating the early stages of construction pricing utilising basic information projects and without the need for more thorough design, according to the data produced from the trained models. Accurate cost assessment is one of the most pressing issues in today's construction projects. However, estimating costs at the start of a project is challenging. Estimators employ a variety of techniques to forecast the cost of construction and civil engineering projects. The goal of this research is to use Artificial Neural Networks to come up with a cost estimate (ANN). Finally, the correctness of these models is determined by a realistic estimation of the procedure, as well as the built Long Short-Term Memory (LSTM) neural networks, as well as developing the neural network model analysis using Microsoft Excel Solver and MATLAB software. Where the parameters (floor height, floor area, total floor area, number of floors) were considered as inputs, (real cost) as target and (estimated cost) as outputs
