Research On The Intellective Audit Of The Key Point Of Government Investment Projects In The Phases Before Construction

People always pay attention to the period from construction to completion when engaging in government investment projects audit in a long time, because of the much money concentrated in the projects construction and some other objective reasons. The finished government investment projects performance audit results show that the government investment projects'efficiency is too low. It is not unrelated to ignoring the process tracking auditing. The government has issued some audit laws that demanding process track auditing to put the axe in the helve. It's a beacon for the government investment audit. But because of the defective theory, lacked experience, and some other objective problems, such as the conflict between the expanding government investment and the lacking capital construction auditor, lagged audit technique methods, the government investment projects process tracking audit is not put in practice materially yet. In some places in China, the government investment audit is only put in practice symbolically. It can't supervise the government investment, but inversely enhance the auditor's risk. So how to put the key point audit in advance to practice is still an important problem to resolve firstly for government investment projects. This paper studies the key point audit in advance in decision making, design and bidding phase by mathematical statistics, FTA method, looking for renting theory, microeconomic theory, neural network ensemble method, and so on.More than 1,000 government investment projects'audit data was analyzed in the paper. We find that most of the government investment projects'performances are not satisfied by these data. Most government investment projects'costs are higher than the average of social lever by statistical data of construction license. The may reason is not the faulty construction, but the imperfect pre-construction. The result indicates that the deregulations occur in many government investment projects'pre-construction phase. It is important to audit in advance in decision-making, design and bidding phase.This paper discusses the feasibility and necessity of application in government investment project audit by neural network ensemble method. The traditional neural network can't minimize the error caused by the different project's complexity of affection factors, experience of design, and the high diversity. This text introduces the neural network ensemble's development and discusses theory of the higher precision. The paper discusses the necessity and content of the decision-making tracking...