【摘要】： 在醫(yī)學(xué),生物,社會(huì)科學(xué)等研究領(lǐng)域,闡明因果關(guān)系已成為許多實(shí)際研究中的最終目的,只有通過(guò)因果關(guān)系才能夠預(yù)見(jiàn)某些行為或策略對(duì)研究對(duì)象的影響。以研究氣陰兩虛癥型和癥狀之間的關(guān)系為例,從醫(yī)學(xué)理論而言,醫(yī)生根據(jù)病人不同的癥狀表現(xiàn)判斷病人是否屬于氣陰兩虛癥型,從而決定相應(yīng)的治療方法�？梢�(jiàn),不同癥狀和氣陰兩虛癥型之間具有因果關(guān)系。但一般的統(tǒng)計(jì)方法如對(duì)數(shù)線性模型雖然可以量化變量之間的關(guān)聯(lián)程度,卻不能體現(xiàn)出這樣一種區(qū)別:已經(jīng)診斷為具有氣陰兩虛癥型的病人具有某一癥狀的可能性,以及具有某一癥狀的病人被診斷為具有氣陰兩虛癥型的可能性。變量間的相關(guān)性是以觀察數(shù)據(jù)為研究對(duì)象,而因果關(guān)系并不完全由數(shù)據(jù)間的相關(guān)性決定,所以對(duì)因果關(guān)系的研究具有不同的意義。 本文將因果圖模型和虛擬事實(shí)模型結(jié)合起來(lái),對(duì)一種簡(jiǎn)單的有向非循環(huán)因果圖建立相應(yīng)的虛擬事實(shí)模型。從而對(duì)醫(yī)學(xué)中關(guān)于癥型和癥狀之間的因果關(guān)系進(jìn)行建模,并且將此模型用于糖尿病患者氣陰兩虛癥型和頭暈癥狀的關(guān)系,以此討論如何將這種模型用于研究在這種病例中氣陰兩虛癥型對(duì)頭暈癥狀的影響程度。
[Abstract]:In the fields of medicine, biology, social science and so on, clarifying causality has become the ultimate goal of many practical researches. Only through causation can we foresee the influence of certain behaviors or strategies on the research object. Taking the study of the relationship between deficiency of qi and yin and symptoms as an example, according to the medical theory, doctors judge whether the patients belong to the syndrome of deficiency of qi and yin according to the different symptoms of the patients, and decide the corresponding treatment methods. It can be seen that there is a causal relationship between different symptoms and deficiency of qi and yin. But general statistical methods, such as logarithmic linear models, which can quantify the degree of correlation between variables, do not reflect the difference: the possibility of a certain symptom in a patient who has been diagnosed as having deficiency of both qi and yin. And the possibility that a patient with a certain symptom is diagnosed with deficiency of Qi and Yin. The correlation between variables is based on observational data, but causality is not entirely determined by the correlation between data, so the study of causality has different meanings. In this paper, the causal graph model and the virtual fact model are combined to establish the corresponding virtual fact model for a simple directed acyclic causality graph. The causal relationship between symptoms and symptoms was modeled in medicine, and the model was applied to the relationship between deficiency of qi and yin and dizziness in patients with diabetes mellitus. This paper discusses how this model can be used to study the influence of deficiency of qi and yin on dizziness in this case.
【學(xué)位授予單位】：華中師范大學(xué)
【學(xué)位級(jí)別】：碩士
【學(xué)位授予年份】：2008
【分類號(hào)】：R311

【參考文獻(xiàn)】

相關(guān)期刊論文 前6條

1 鄭忠國(guó),童行偉,張艷艷;因果模型中因果效應(yīng)的可識(shí)別性研究[J];中國(guó)科學(xué)(A輯);2001年12期

2 耿直,何洋波,王學(xué)麗;因果鏈上因果效應(yīng)的關(guān)系及推斷[J];中國(guó)科學(xué)(A輯:數(shù)學(xué));2004年02期

3 趙慧;鄭忠國(guó);;圖模型及其在統(tǒng)計(jì)建模中的應(yīng)用[J];數(shù)理統(tǒng)計(jì)與管理;2007年03期

4 溫晗秋子;占座效應(yīng)的數(shù)學(xué)建模與統(tǒng)計(jì)分析[J];數(shù)學(xué)的實(shí)踐與認(rèn)識(shí);2003年10期

5 易丹輝;王為;;對(duì)數(shù)線性模型在中醫(yī)數(shù)據(jù)中的應(yīng)用[J];統(tǒng)計(jì)與決策;2006年15期

6 耿直;因果推斷與Simpson悖論[J];統(tǒng)計(jì)與信息論壇;2000年03期

，

本文編號(hào)：2142648