会计类英语翻译3（很急）
急！急！非常急！麻烦各位英语牛人帮忙啊！！！翻译下面这几段的会计类的英语，在线等，在线等，急用！
要求： 要比较专业的翻译，不要用机器翻译的，谢谢啦~~~~~~
The solution of the problem was the application of other statistical methods - so in 1980 Ohlson (1980) developed a conditional logistic regression model. This logit bankruptcy prediction model allowed for the estimation of the probability of bankruptcy conditional on the values of nine financial ratios and enabled the subsequent studies to make more detailed analyses of company's financial performance and to establish cause-and-effect relations with respect to the industrial sector, company size, etc.
After Ohlson, an innovation in the prognostic models was introduced in 1987 by Healy (1987), who used the time series methodology (multivariate cumulative sum - the CUSUM method). The advantage of time series models application in predicting company's financial performance is that this method takes into account the correlation of data within a time series. In explanation, these models distinguish between the changes of financial ratios that occur as a result of the correlation data and the changes that occur in the structure of financial ratios as a result of bad financial performance. At the same time, according to Theodossiou (1993:448) "The CUSUM model can be viewed as the dynamic extension of discriminant analysis, a statistical technique used in many business failure prediction studies."
The next turn in the development of financial performance prediction models arouse with the introduction of human behavior simulation models. For instance, in 1994 Altaian, Marco and Varetto (Altman and Hotchkiss, 2006) combined linear discriminant analysis and neural networks in their research on the sample of 37.000 small and medium-sized companies. They applied the linear discriminant analysis in two steps to refine the model, and the results obtained were repeated in neural network model that is more flexible and provides more precise solutions, especially for specific complex cases.
In addition to human behavior simulation models, a number of other statistical techniques such as probit models, decision tree analysis or Bayesian discriminant analysis are used today.
奥尔森之后，1987年Healy引入了一个全新的预测模型。他使用了时间序列方法论（多元累积总合 - CUSUM法）。使用时间序列模式来预测公司的金融状况的优势是这个方法将一段时间之内的数据相互关系考虑在内.这些模型区分相关数据引发的金融比值变化和由恶劣的金融状况引起的金融比值结构变化。同时，根据Theodossiou “CIUSIUM模型可以被看作判别式分析的动态延伸，一个被运用在许多商业失败预测研究中的统计技巧。”
金融状况预测模型的下一个转折点是人类行为刺激模型。1994， Altian， Marco和Varette(Altman and Hotchkiss, 2006)在他们的关于37000个中小型公司的研究中综合了线性判别分析和神经网络。他们在两个步骤里使用线性判别分析来完善模型，得到的结果在神经网络模型里再次重复。神经网络模型更具弹性，提供更精确的解法，尤其是对于特定的复杂案例。
除了人类行为刺激模型，一些其它的统计技巧例如概率单位模型，决定树分析和贝叶斯判别分析也在今天被使用。
