Prediction And Development Of Total Output Value Of Construction Industry Based On Machine Learning

The construction industry has an important position and role in the development of China’s national economy,and it is important to achieve quantitative analysis and scientific forecasting of the scale of construction output value.However,most of the existing construction economic forecasting methods have low accuracy and ignore the changes of relevant influencing factors,which has certain limitations.Therefore,this paper introduces machine learning theory into the research on the prediction and development of the total construction industry output value,based on random forest,BP neural network and support vector machine algorithms,and takes the construction industry in Hubei Province as an example,through index screening,model construction and optimisation,to carry out empirical analysis and prediction of the total construction industry output value in Hubei Province.Firstly,a theoretical analysis of the research results of construction industry development and economic growth laws was conducted to summarise 17 factors related to the gross output value of the construction industry and construct an index system of factors influencing the gross output value of the construction industry.Based on the random forest algorithm,10 key factors influencing the gross output value of the construction industry,including gross regional product,the number of survey and design institutions,total energy consumption in the construction industry and the number of enterprise units,were obtained,laying the foundation for the establishment of the construction industry gross output value prediction model.Secondly,this paper takes the screened key influencing factors of gross output value as input variables,constructs a BP neural network and GSM-optimised SVM forecasting model,uses the data set of gross output value and related influencing factors of the construction industry in Hubei Province as training samples,conducts simulation and carries out systematic comparison and analysis with the existing models GSM-SVM1 and BPNN1 and the traditional ARIMA model.The results show that both the GSM-SVM2 and BPNN2 forecasting models proposed in this paper can effectively predict the scale and growth trend of the gross construction output value,verifying the feasibility and practicality of machine learning theory in the field of gross construction output value forecasting.In addition,the constructed models have higher forecasting accuracy than existing forecasting models and traditional models,with the GSM-SVM2 model’s forecasting results more closely resembling the actual values,resulting in better overall forecasting results.Again,to address the problem of selecting the core parameters C and g in the support vector machine,a particle swarm optimisation algorithm was used to optimise the parameters of the support vector machine model and construct a combined PSO-SVM prediction model.Through the performance test and comparative analysis of the model,the PSO-SVM forecasting model with the best forecasting effect was selected to forecast and analyse the total construction output value in Hubei Province from 2021 to 2025.The results show that the optimised PSO-SVM forecasting model has higher solution accuracy and generalisation capability,and is a more effective method for modelling and forecasting the gross construction output value.In addition,in the next five years,the total construction output value of Hubei Province will always maintain the growth trend and the industry scale will be further increased,and it is expected to achieve the development goal of reaching a total construction output value of over RMB 2.5 trillion in 2025.Finally,based on the forecast and analysis results and combined with macroeconomic policies,measures and suggestions related to industry development such as promoting informatization,strengthening talent training and developing international markets are proposed for the construction industry in Hubei Province,aiming to provide certain information support for the relevant departments to formulate policies and norms for the economic development of the construction industry and promote the growth and long-term prosperity of the total construction output value.