Research On The Investment Prediction And Control System About The STC Real Estate Project Of L Group

In recent years,the profitability of real estate enterprises has generally declined,the overall profit scale has shown negative growth,and the profit margin has suffered a "Waterloo type" flash collapse and losses,and even to the verge of bankruptcy.L Group,which mainly focuses on real estate development,also faces the same problem.The investment control of the boutique residential projects it develops is facing embarrassment.At the completion and settlement,it often has to increase the project money of 10% of the budget,and the investment overspend is serious.In order to improve the investment management method and effectively solve the problem of investment overspending,this thesis constructed the intelligent investment management system(including intelligent prediction of investment objectives and investment control dynamic early warning optimization),and applied it to STC-R area project,try to break through the disadvantages of traditional quota prediction,provide more accurate and feasible investment target prediction method,try to process control the project through dynamic early warning optimization,to ensure the realization of investment objectives.The construction process and application results of the system are as follows:First,the historical investment data of the completed projects of L Group are deeply analyzed,and the case database is established as the data basis for the project investment target prediction model of STC-R area projects.Select the cases with high similarity to the project according to the cosine similarity principle,and select the appropriate investment target prediction model according to the number of similar cases(i.e.,use neural network model for more than 10 similar cases,and use case inference model for 1-10 similar cases).Secondly,an appropriate prediction model is used to predict the total investment target,total labor cost,material cost and construction usage of a residential building in the R area;according to the characteristics of the project,determine the investment target of each sub-purpose according to the investment proportion of each sub-purpose.The results show that the average relative error of the neural network prediction model is 2.73%,less than 3%,which is better than the conventional set of quota pattern.Again,through collecting L group has completed the investment overspend problems,influencing factors and countermeasures to establish investment control countermeasure database,based on the earn value method to establish investment control early warning system,real-time monitoring of dynamic investment project,to help managers for process control,in the case of overspend warning quickly adopt appropriate solution strategy.After the application of the system,the response speed of handling the problems is faster,and the control efficiency of the investment overspend problem is improved,up by 9.2% than before not using the system.Finally,according to the PDCA management cycle model,dynamic monitoring and real-time adjustment strategies are established in PDCA.Then,the planned investment and actual investment of each subsidiary are compared in the cycle cycle,the cumulative target value and actual value are compared,and problems are found in time and deviations are corrected in real time.The results show that the actual investment of a single residential project in Zone R saves 580,000 yuan compared with the target investment,increases the enterprise profit,and provides technical support and control scheme for the survival of enterprises under the current tightening external situation.