T*******I 发帖数: 5138 | 1 我打算今年去JSM上讲讲这个新概念系统。我不是学数学的,所以提出来请大家帮忙修
正一下,免得到时候闹笑话。当然,到时候是全英文版的。估计版上有些人已经看过的。
还会有一些数学符号式的表述,这里无法显示。多谢了。
个体:在认识论范畴内,一个个体是一个独立的存在或实体或客体,且拥有其自身已知
的、可知的和不可知的全部属性,并且由于这些属性,一个个体可以与所有其它个体相
区别。在一个特定的领域中,任何以最小单元存在着的事物可以被称为是一个个体。当
一个个体进入一个主体的观察范畴且能被认知或再认知时,它的每一个属性应该是确定
的而非不确定的。换句话说,一个个体是它自己而非任何其它事物是由于它所拥有的全
部属性至少在被认知的那一刻是确定的。反之,如果它的全部属性在被主体观察时是不
确定的,那么主体将对它不可知,或者说它对于主体来说不可测。
属性:一个个体的一个属性(用符号A(字体:kunstler script)表示)是关于它的一个
抽象的特征。这类抽象的特征通常有质和量两大类,由此我们可以在许多个体中定义一
个群体或类。例如,一个个体可以有姓字、性别、身高和体重等属性。每一个属性是唯
一的并且表达着一个特定的含义。
子属性:它是一个附属的属性且被定义在一个属性的名下,例如,姓名={亚里士多德,
培根,黑格尔},性别={男,女,性别畸变}以及年龄={介于[0,140]之间的一个数值,
如2,35或86岁},等等,其中{亚里士多德,培根,黑格尔}、{男,女,性别畸变}和{2
, 35 or 86岁}等是被分别定义在姓名、性别和年龄等名下的子属性。
不变属性:一个属性被认为是不变的,如果(1)它是它自己;或(2)没有子属性可以被定
义在其名下;或(3)即使存在子属性,但定义它们是不必要的。从而,这个属性在观察
或试验过程中可以被认为是没有变化或变异性的,因而可以被用来清楚地定义一个群体
或类别,例如,性别=男,或年龄大于等于18岁,或性别=男且年龄大于等于18岁,等等。
可变属性:一个属性被认为是可变的,如果在一个观察或试验中至少有两个不同的子属
性可以被定义在其名下,且各子属性是可以相互区分且准确定义的,相互之间没有任何
混淆和冲突。因此,可变属性的概念等同于现行系统中随机变量的概念,例如,性别={
男,女,性别畸变},0岁<=年龄<=140岁,等。
离散可变属性:一个属性被认为是离散可变的,如果定义在其名下的所有子属性是质性
的,例如,地点和学校、树木和湖泊、疾病和治疗,等等。
连续可变属性:一个属性被认为是连续可变的,如果定义在其名下的所有子属性是量性
的,例如,高度和重量、速度和加速度、容积和比率,等等。
总体或总体空间:一个总体(用符号P(字体:kunstler script))是由一些有着相同的不
变和可变属性的个体组成的一个群体或集合。总体中的个体构成了一个空间,即总体空
间。通常,一个总体被认为是无限的,因为其中的个体数量可能是无限的,或者由于数
量巨大以至于在一次有限的观察中不可能全部观察到。一个总体有可能进入一个或一群
观察主体的一个特定的观察或试验范畴。
尺度空间:一个尺度空间(用符号Ω表示)是由一个可变属性的全部无重复或冲突的子属
性或一次观察或试验中的全部可能结果构成的空间,例如,一个统计调查表就是一个尺
度空间。由此,一个尺度空间是关于不变属性和可变属性的一个集合,且这个集合不能
为空集,因为它是一个统计测量的工具。因此,这里对尺度空间的定义等同于现行概率
论系统中的“样本空间”的定义。显然,一个尺度空间不能被说成是一个样本空间,因
为它仅仅是一个测量工具而非一个样本本身。
测度:一个测度(用符号M表示)是在一定的观察或试验范畴内有着特定目的的测量行为
,通常由至少一个主体执行以便获得关于总体中一定数量的个体的不变属性和可变属性
的原始记录和认知。特别地,在统计学中的所有测度都是随机测度,因为任何被测对象
都是随机得到的。
分布:一个分布(用符号D表示)是关于个体的观察结果在尺度空间上的表达。
样本:一个样本(用符号S表示)是一个测量行为中全部被观察个体的全部结果,因此,
它是一个尺度空间上完整的分布。一个样本是总体的一个随机子集。不存在没有尺度空
间相关联的独立样本;反之亦然。在统计学范畴内,一个样本通常也被称为是一个数据
集。由于总体中个体的无限性,一个样本应该通过一个随机机制获得从而使得其对总体
的代表性得到一定程度的保证。由此,统计学范畴内的任何样本都是一个随机样本。在
统计学中,样本中的一个个体通常被称为一个“观察”或“随机样本点”或简称“样本
点”,因此,一个样本中一个个体或观察或样本点不能再被称为是一个“样本”;否则
将引起概念间的混淆甚至冲突;除非在一个测量行为中只有一个个体被观察到,此时,
一个样本就等于这个个体。一般而言,一个样本自身作为一个整体在另一个观察范畴内
是一个个体,但却是不同于样本中的个体的个体。这个作为“个体”的样本也应该拥有
其自身的属性,即样本属性,且每一个属性也应该是确定的,恰如以上讨论的关于总体
中个体的属性的性质一样。
样本空间:一个样本空间(表示符号同样本)可以是一个样本自身或样本数据集,因为在
任何样本中应该没有重复的个体记录,因而每个样本点都是一个独立的元素,即使在仅
有一个离散变量而关于该变量的观察仅有两个或两个以上的子属性和三个或三个以上的
观察个体的下情形中也是如此。换句话说,我们可以反问:如果一个样本自身不能被称
为是一个样本空间,那么,还有什么其它的东西能被称为是样本空间呢?事实上,一个
样本中的全部个体就构成了一个完整的空间,这个空间就是样本空间。
可测空间:一个空间被认为是可测的,如果其中每一个体在尺度空间上可测。从而,一
个总体是一个可测空间,因为其中所有的个体在一个尺度空间上应该是可测的。
被测空间:一个空间被认为是被测的,如果其中每一个体被一个尺度空间所测,无论这
个测量对于任一个体是否成功。从而,一个样本是一个被测空间。
随机映射:它是一个随机机制,用符号M(字体:kunstler script)。通过它一个样本或
样本空间被从一个可测空间或总体在尺度空间上得到。
概率空间:一个概率空间(用符号P表示)是一个被概率化为1的样本空间。我们不能将一
个概率空间定义在一个总体空间或可测空间上,因为一个尺度空间对于一个总体来说可
能不是一个完备的空间但对于样本来说却是完备的。此外,一个总体空间通常是未知的
,因此,一个概率空间如果被定义在一个总体空间上将带给我们一个未知的空间,从而
这样的定义是徒劳的。我们也不能将概率空间单独地定义在一个尺度空间上,因为后者
不过是一个测量工具而非我们试图通过概率来认识的真实的随机世界。然而,一个概率
空间应该是被定义在一个分布着样本空间中的全部被测个体的尺度空间上。因此,只有
样本空间是一个完备的空间且可以在尺度空间上被概率化。当然,一个在数学上被很好
地定义了的确定的完备空间也是可以被概率化为1的,只要它满足由当前知识系统设定
的一些特定的条件,例如,所有理论分布,包括正态分布、标准正态分布、t-分布、F-
分布以及卡方分布,等等。因此,如何概率化一个样本空间属于数学特别是概率论的范
畴。
空间的连续性和可连续性:由于总体的无限性,我们不能在总体空间上直接讨论空间的
连续性,但可以经由样本来讨论这个问题。这里有两个不同的概念:一个是连续空间;
另一个是可连续空间。一个连续空间不等于一个可连续空间。一个样本空间被认为是连
续的,如果其中所有个体处于一个确定的子样本空间或整个样本空间自身之中,例如,
100个男性的身高和100个女性的身高将各自被视为一个连续空间而不是一个可连续空间
。然而,如果将这两个空间混合在一起,则这200人的身高将被视为是一个可连续空间
而非一个连续空间,因为这个混合空间是由两个可识别的、相互重叠或分离的空间构成
的。不过,这个混合空间仍然可以以一种连续测量的方式得到,且以“人的身高”为属
性被定义为一个连续空间。
空间的不可分性和可分性:一个空间的可分性在离散空间里是很容易理解的。曾经令人
在哲学上感到困难的是关于一个连续空间的可分性,例如,一块砖头是一个完整的连续
空间,如果将它分开,势必要打破它。然而,在引入了空间的可连续性概念后,这样的
理解就不会遇到任何逻辑障碍,例如,由两块砖头粘合起来的空间也可以被视为一个可
连续空间,但却一个可分离的空间,因此,一个可连续空间不等于一个连续空间,而一
个可连续空间具有可分性。
统计量:一个统计量(用符号s表示)是关于样本或样本空间的一个属性。由于样本是来
自总体的一个随机子集,因此,一个统计量是一个随机的点测量。它也被认为是定义在
样本空间也就是概率空间上的一个实可测函数。统计学所要做的正是构造特定的统计量
以便对样本空间的属性作出描述,进而推断总体空间的相应属性(总体的属性通常用一
个特定的术语即参数来称呼)。因此,一个统计量是一个随机的常量而非一般数学意义
上的常量。一般数学意义上的常量通常没有任何形容词修饰,也就是一个常量是它自己
。由此可知,一个样本中的全部记录也都可以被理解为随机常量。在统计学的范畴内,
一个常量被认为是随机的仅仅是针对样本本身而言。因此,我们可以说一个统计量对于
一个给定的样本来说是确定的,而对于总体来说则是非确定的。然而,一个样本统计量
在不同的子样本之间以及它们与整个样本的统计量之间可以是不同的,因为任何子样本
为其自身的统计量提供了较少的信息。例如,一个单一的全域回归模型提供了关于整个
样本空间上的一个确定不变的回归关系,而分段回归模型将带给我们一组不同临界空间
上的可变关系,从而,一个完整的样本空间可以被分割为若干个片段。
参数:一个参数(用符号p表示)是关于总体的一个属性,通常用一个相应的样本统计量
来估计和推断,此时的总体参数可以被认为是不变的,且这一假设对总体来说不会导致
损害。然而,我们必须意识到它在自身的自然历史中应该是可变的。
随机空间或随机系统:在我们所讨论的问题的范畴内,一个随机空间或随机系统(用符
号R(字体:kunstler script)表示)是一个与上述全部概念相关联的抽象概念,也就是
说,它是一个广义化的概念,而非上面提到的某个或某几个具体的概念。由于定义总体
的不变属性和样本中个体的随机常量以及样本本身的全部统计量,一个随机空间可能包
含了一定程度的确定性,从而在描述和推断总体时,我们的结论也就有了一定程度的确
定性。但是,我们必须牢记,任何样本对其总体的非确定性是一个绝对的本质属性,因
此,基于样本基础上的关于总体的全部描述在本质上是随机的或非确定性的。 |
c****r 发帖数: 969 | 2 i am sure ppl will fall sleep.
的。
【在 T*******I 的大作中提到】 : 我打算今年去JSM上讲讲这个新概念系统。我不是学数学的,所以提出来请大家帮忙修 : 正一下,免得到时候闹笑话。当然,到时候是全英文版的。估计版上有些人已经看过的。 : 还会有一些数学符号式的表述,这里无法显示。多谢了。 : 个体:在认识论范畴内,一个个体是一个独立的存在或实体或客体,且拥有其自身已知 : 的、可知的和不可知的全部属性,并且由于这些属性,一个个体可以与所有其它个体相 : 区别。在一个特定的领域中,任何以最小单元存在着的事物可以被称为是一个个体。当 : 一个个体进入一个主体的观察范畴且能被认知或再认知时,它的每一个属性应该是确定 : 的而非不确定的。换句话说,一个个体是它自己而非任何其它事物是由于它所拥有的全 : 部属性至少在被认知的那一刻是确定的。反之,如果它的全部属性在被主体观察时是不 : 确定的,那么主体将对它不可知,或者说它对于主体来说不可测。
|
s*****r 发帖数: 790 | 3 You have shown how naive you are about all the concepts here. You have
absolute no understanding of the terminologies you used, except the words
linguistically.
for example, 总体中的个体构成了一个空间,即总体空间。Why this is a space? Do
you know for a set to be a space what properties it should have? do you
know the definition of "space"?
I bet you know nothing about the concepts of countable and uncountable.
btw, your "measure", "measurable", to me sounds like "observation" or
observable.
是学数学的,所以提出来请大家帮忙修正一下,免得到时候闹笑话。当然,到时候是全
英文版的。估计版上有些人已经看过的。还会有一些数学符号式的表述,这里无法显示
。多谢了。
知的、可知的和不可知的全部属性,并且由于这些属性,一个个体可以与所有其它个体
相区别。在一个特定的领域中,任何以最小单元存在着的事物可以被称为是一个个体。
当一个个体进入一个主体的观察范畴且能被认知或再认知时,它的每一个属性应该是确
定的而非不确定的。换句话说,一个个体是它自己而非任何其它事物是由于它所拥有的
全部属性至少在被认知的那一刻是确定的。反之,如果它的全部属性在被主体观察时是
不确定的,那么主体将对它不可知,或者说它对于: 魈謇此挡豢刹狻
个抽象的特征。这类抽象的特征通常有质和量两大类,由此我们可以在许多个体中定义
一个群体或类。例如,一个个体可以有姓字、性别、身高和体重等属性。每一个属性是
唯一的并且表达着一个特定的含义。
,培根,黑格尔},性别={男,女,性别畸变}以及年龄={介于[0,140]之间的一个数值
,如2,35或86岁},等等,其中{亚里士多德,培根,黑格尔}、{男,女,性别畸变}和
{2, 35 or 86岁}等是被分别定义在姓名、性别和年龄等名下的子属性。
定义在其名下;或(3)即使存在子属性,但定义它们是不必要的。从而,这个属性在观
察或试验过程中可以被认为是没有变化或变异性的,因而可以被用来清楚地定义一个群
体或类别,例如,性别=男,或年龄大于等于18岁,或性别=男且年龄大于等于18岁,等
等。
属性可以被定义在其名下,且各子属性是可以相互区分且准确定义的,相互之间没有任
何混淆和冲突。因此,可变属性的概念等同于现行系统中随机变量的概念,例如,性别
={男,女,性别畸变},0岁<=年龄<=140岁,等。
性的,例如,地点和学校、树木和湖泊、疾病和治疗,等等。
性的,例如,高度和重量、速度和加速度、容积和比率,等等。
不变和可变属性的个体组成的一个群体或集合。总体中的个体构成了一个空间,即总体
空间。通常,一个总体被认为是无限的,因为其中的个体数量可能是无限的,或者由于
数量巨大以至于在一次有限的观察中不可能全部观察到。一个总体有可能进入一个或一
群观察主体的一个特定的观察或试验范畴。
【在 T*******I 的大作中提到】 : 我打算今年去JSM上讲讲这个新概念系统。我不是学数学的,所以提出来请大家帮忙修 : 正一下,免得到时候闹笑话。当然,到时候是全英文版的。估计版上有些人已经看过的。 : 还会有一些数学符号式的表述,这里无法显示。多谢了。 : 个体:在认识论范畴内,一个个体是一个独立的存在或实体或客体,且拥有其自身已知 : 的、可知的和不可知的全部属性,并且由于这些属性,一个个体可以与所有其它个体相 : 区别。在一个特定的领域中,任何以最小单元存在着的事物可以被称为是一个个体。当 : 一个个体进入一个主体的观察范畴且能被认知或再认知时,它的每一个属性应该是确定 : 的而非不确定的。换句话说,一个个体是它自己而非任何其它事物是由于它所拥有的全 : 部属性至少在被认知的那一刻是确定的。反之,如果它的全部属性在被主体观察时是不 : 确定的,那么主体将对它不可知,或者说它对于主体来说不可测。
|
T*******I 发帖数: 5138 | 4 You mean that the "definition of space" JUST in Mathematics? What if about I
define a concept of "space" in a Non-mathematical environement? Do you
really think that there must be a logical obstacle to you to understand the
"population space" that I definition here? If yes, I might make a mistake.
So I need your help to clarify it. Thanks very much!
Do
【在 s*****r 的大作中提到】 : You have shown how naive you are about all the concepts here. You have : absolute no understanding of the terminologies you used, except the words : linguistically. : for example, 总体中的个体构成了一个空间,即总体空间。Why this is a space? Do : you know for a set to be a space what properties it should have? do you : know the definition of "space"? : I bet you know nothing about the concepts of countable and uncountable. : btw, your "measure", "measurable", to me sounds like "observation" or : observable. :
|
A*******s 发帖数: 3942 | 5 as i suggested before, call it Chen space. Similarly, Chen measurement, Chen
stochastic process, Chen random variable, etc...
I
the
【在 T*******I 的大作中提到】 : You mean that the "definition of space" JUST in Mathematics? What if about I : define a concept of "space" in a Non-mathematical environement? Do you : really think that there must be a logical obstacle to you to understand the : "population space" that I definition here? If yes, I might make a mistake. : So I need your help to clarify it. Thanks very much! : : Do
|
s*****r 发帖数: 790 | 6 don't call it a space, it is just a set.
if you don't want people to think those terms in math, use your know terms,
like Actuaries suggested.
please don't use any mathematical terms, since you don't understand them.
you don't need to learn since you won't be able to understand anyway.
so just do NOT use them.
I
the
【在 T*******I 的大作中提到】 : You mean that the "definition of space" JUST in Mathematics? What if about I : define a concept of "space" in a Non-mathematical environement? Do you : really think that there must be a logical obstacle to you to understand the : "population space" that I definition here? If yes, I might make a mistake. : So I need your help to clarify it. Thanks very much! : : Do
|
T*******I 发帖数: 5138 | 7 So, let me try to get an answer: What is a space in Mathematics?
,
【在 s*****r 的大作中提到】 : don't call it a space, it is just a set. : if you don't want people to think those terms in math, use your know terms, : like Actuaries suggested. : please don't use any mathematical terms, since you don't understand them. : you don't need to learn since you won't be able to understand anyway. : so just do NOT use them. : : I : the
|
s*****r 发帖数: 790 | 8 this is a good start: you want to learn.
you can google it, such as http://en.wikipedia.org/wiki/Space_%28mathematics%29
【在 T*******I 的大作中提到】 : So, let me try to get an answer: What is a space in Mathematics? : : ,
|
T*******I 发帖数: 5138 | 9 That is good. I will take a look.
One more question: Is ther any other issues in my conceptual system except
for the "space" issue?
Thanks very much.
【在 s*****r 的大作中提到】 : this is a good start: you want to learn. : you can google it, such as http://en.wikipedia.org/wiki/Space_%28mathematics%29
|
s*****r 发帖数: 790 | 10 you should ask: is there any concept in my conceptual system that does not
have an issue.
【在 T*******I 的大作中提到】 : That is good. I will take a look. : One more question: Is ther any other issues in my conceptual system except : for the "space" issue? : Thanks very much.
|
|
|
T*******I 发帖数: 5138 | 11 Ok, you can make your claim or criticism on any aspect that you think that
might not be correct. Thanks. Sincerely.
【在 s*****r 的大作中提到】 : you should ask: is there any concept in my conceptual system that does not : have an issue.
|
s*****r 发帖数: 790 | 12 treat it as a paper, if I am the reviewer, I will likely just say it is not
suitable for publication and will not provide a detailed comments.
Only those who are so free that their balls are itching will give you a
detailed comments, basically, educating you math 101.
If you are serious about this, I suggest you take a peek at some classic
math book and find out what terms show up in your "paper", then figure out
what those terms mean, not just from a dictionary and you think.
【在 T*******I 的大作中提到】 : Ok, you can make your claim or criticism on any aspect that you think that : might not be correct. Thanks. Sincerely.
|
g********r 发帖数: 8017 | 13 你真有耐心。这些话版上的人重复了几十遍了。
其实一个统计本科的课程,除掉乱七八糟的马列,体育之类的,自学两年够了。大师不
肯。
not
【在 s*****r 的大作中提到】 : treat it as a paper, if I am the reviewer, I will likely just say it is not : suitable for publication and will not provide a detailed comments. : Only those who are so free that their balls are itching will give you a : detailed comments, basically, educating you math 101. : If you are serious about this, I suggest you take a peek at some classic : math book and find out what terms show up in your "paper", then figure out : what those terms mean, not just from a dictionary and you think.
|
T*******I 发帖数: 5138 | 14 However, you have agreed with me at
Statistics =/= Mathematics
and I don't think Statistics is just a branch of mathematics. So, why I must learn math though it does discuss Statistics in its own style? I would like to try another style. Why not? Did I really make mistakes from the angle of mathematics? Did I really make mistakes from the angle of Statistics?
In my opinion, Statistics should emerge before mathematics.
【在 s*****r 的大作中提到】 : treat it as a paper, if I am the reviewer, I will likely just say it is not : suitable for publication and will not provide a detailed comments. : Only those who are so free that their balls are itching will give you a : detailed comments, basically, educating you math 101. : If you are serious about this, I suggest you take a peek at some classic : math book and find out what terms show up in your "paper", then figure out : what those terms mean, not just from a dictionary and you think.
|
y**t 发帖数: 205 | 15 You know when you try to present your ideas, to make communication easier,
you want to talk in the same language as majority of your audiences do.
Your problem is that you use several stat terms but not in their correct
definition. This will cause huge trouble when communicate with people who
toke the training in statistics or mathematics.
As others already suggested, you'd better 1) define your own terms and avoid
the well-known name in stat text book or 2) used the existed term in the
right way.
Consider your work, I suggest 1).
must learn math though it does discuss Statistics in it own style? I would
like to try another style. Why not? Did I really make mistakes from the
angle of mathematics? Did I really make mistakes from the angle of
Statistics?
【在 T*******I 的大作中提到】 : However, you have agreed with me at : Statistics =/= Mathematics : and I don't think Statistics is just a branch of mathematics. So, why I must learn math though it does discuss Statistics in its own style? I would like to try another style. Why not? Did I really make mistakes from the angle of mathematics? Did I really make mistakes from the angle of Statistics? : In my opinion, Statistics should emerge before mathematics.
|
T*******I 发帖数: 5138 | 16 既然它们被定义好了,为什么在应用统计领域从来不讲授那些数学里定义好了的概念:
诸如测度空间、可测性、测度分布、样本空间等等。WHY? Who can talk these
concepts clearly without any confusion to a non-mathematical-background
statistician ? Can you?
统计学家们面对的是真实的世界,而这个真实的世界不是从数学理论中演绎出来的,而
是恰恰相反!因此,没有人会care你的数学理论是如何的,人们关心的是如何更好地认
识真实的世界。
更何况并非只有数学背景的人才能从事统计行当。人们需要一个简洁直观的概念系统来
帮助他们一边工作一边思考。
请不要继续蒙人了。
avoid
【在 y**t 的大作中提到】 : You know when you try to present your ideas, to make communication easier, : you want to talk in the same language as majority of your audiences do. : Your problem is that you use several stat terms but not in their correct : definition. This will cause huge trouble when communicate with people who : toke the training in statistics or mathematics. : As others already suggested, you'd better 1) define your own terms and avoid : the well-known name in stat text book or 2) used the existed term in the : right way. : Consider your work, I suggest 1). :
|
A*******s 发帖数: 3942 | 17 when u use computer do u need to know quantum mechanics?
when u use electricity from nuclear plant do u need to know relativity
theory?
【在 T*******I 的大作中提到】 : 既然它们被定义好了,为什么在应用统计领域从来不讲授那些数学里定义好了的概念: : 诸如测度空间、可测性、测度分布、样本空间等等。WHY? Who can talk these : concepts clearly without any confusion to a non-mathematical-background : statistician ? Can you? : 统计学家们面对的是真实的世界,而这个真实的世界不是从数学理论中演绎出来的,而 : 是恰恰相反!因此,没有人会care你的数学理论是如何的,人们关心的是如何更好地认 : 识真实的世界。 : 更何况并非只有数学背景的人才能从事统计行当。人们需要一个简洁直观的概念系统来 : 帮助他们一边工作一边思考。 : 请不要继续蒙人了。
|
T*******I 发帖数: 5138 | 18 如此苍白的反驳。
【在 A*******s 的大作中提到】 : when u use computer do u need to know quantum mechanics? : when u use electricity from nuclear plant do u need to know relativity : theory?
|
s*****r 发帖数: 790 | 19 Your "theory" is about random variable, right? defining a random variable is
not just saying the variable takes different values. If you are serious,
tell us how to define a random variable, either you know from classic
statistics, or your new theory.
if you don't understand the definition of rv, how could you develop a theory
about it?
must learn math though it does discuss Statistics in it own style? I would
like to try another style. Why not? Did I really make mistakes from the
angle of mathematics? Did I really make mistakes from the angle of
Statistics?
【在 T*******I 的大作中提到】 : However, you have agreed with me at : Statistics =/= Mathematics : and I don't think Statistics is just a branch of mathematics. So, why I must learn math though it does discuss Statistics in its own style? I would like to try another style. Why not? Did I really make mistakes from the angle of mathematics? Did I really make mistakes from the angle of Statistics? : In my opinion, Statistics should emerge before mathematics.
|
A*******s 发帖数: 3942 | 20 "我只管用我自己定义的术语来阐述我的新理论,至于数学脑袋们懂不懂,那是他们的
问题,与我无关。"
is
theory
【在 s*****r 的大作中提到】 : Your "theory" is about random variable, right? defining a random variable is : not just saying the variable takes different values. If you are serious, : tell us how to define a random variable, either you know from classic : statistics, or your new theory. : if you don't understand the definition of rv, how could you develop a theory : about it? : : must learn math though it does discuss Statistics in it own style? I would : like to try another style. Why not? Did I really make mistakes from the : angle of mathematics? Did I really make mistakes from the angle of
|
|
|
b*********n 发帖数: 2975 | 21 那个是哲学,你们不懂的, 呵呵
Your "theory" is about random variable, right? defining a random variable is
not just saying the variable takes different values. If you are serious,
tell us how to define a random variable, either you know from classic
statistics, or your new theory.
if you don't understand the definition of rv, how could you develop a theory
about it?
must learn math though it does discuss Statistics in it own style? I would
like to try another style. Why not? Did I really make mistakes from the
angle of mathematics? Did I really make mistakes from the angle of
Statistics?
【在 s*****r 的大作中提到】 : Your "theory" is about random variable, right? defining a random variable is : not just saying the variable takes different values. If you are serious, : tell us how to define a random variable, either you know from classic : statistics, or your new theory. : if you don't understand the definition of rv, how could you develop a theory : about it? : : must learn math though it does discuss Statistics in it own style? I would : like to try another style. Why not? Did I really make mistakes from the : angle of mathematics? Did I really make mistakes from the angle of
|
a***g 发帖数: 2761 | 22 itz so cool
Chen
【在 A*******s 的大作中提到】 : as i suggested before, call it Chen space. Similarly, Chen measurement, Chen : stochastic process, Chen random variable, etc... : : I : the
|
y**t 发帖数: 205 | 23 I was providing my comments since you asked for it. You can take it or not
and it's totally depends on you.
Just to answer your question, I was TA for the measure theory when I was in
graduate school. I DID explain the concept you mentioned to some students
from biology department and they DO can understand them correctly.
【在 T*******I 的大作中提到】 : 既然它们被定义好了,为什么在应用统计领域从来不讲授那些数学里定义好了的概念: : 诸如测度空间、可测性、测度分布、样本空间等等。WHY? Who can talk these : concepts clearly without any confusion to a non-mathematical-background : statistician ? Can you? : 统计学家们面对的是真实的世界,而这个真实的世界不是从数学理论中演绎出来的,而 : 是恰恰相反!因此,没有人会care你的数学理论是如何的,人们关心的是如何更好地认 : 识真实的世界。 : 更何况并非只有数学背景的人才能从事统计行当。人们需要一个简洁直观的概念系统来 : 帮助他们一边工作一边思考。 : 请不要继续蒙人了。
|
T*******I 发帖数: 5138 | 24 If what you said is true, you did a really great job. I wish I could be one
of the biological students. I must appreciate your job. However, I have
never had such a chance.
So, let me ask you, what is a difference between a measure in Math and in a
real activity of measurement? suppose I was one of your student.
in
【在 y**t 的大作中提到】 : I was providing my comments since you asked for it. You can take it or not : and it's totally depends on you. : Just to answer your question, I was TA for the measure theory when I was in : graduate school. I DID explain the concept you mentioned to some students : from biology department and they DO can understand them correctly.
|
s*****r 发帖数: 790 | 25 what is "a real activity of measurement"? can you define it? without
definition nobody can answer your question. remember, examples are not
definitions.
one
a
【在 T*******I 的大作中提到】 : If what you said is true, you did a really great job. I wish I could be one : of the biological students. I must appreciate your job. However, I have : never had such a chance. : So, let me ask you, what is a difference between a measure in Math and in a : real activity of measurement? suppose I was one of your student. : : in
|
T*******I 发帖数: 5138 | 26 我想我并非只是关心如何定义随机变量,而是定义那些我认为在统计学中最基础的几个
概念。这些概念是从人类最一般的统计行为中抽象出来的。这些行为包括了一般的抽样
调查和统计量的构造等。这些活动不是数学要关心的问题,因为数学不是一门研究统计
行为模式的学科,而是一门关于与“数学”有关的抽象概念的逻辑推理的学科。
我在这里组织的这些概念即使在非统计的人来看也是可以理解的。你们也可以感觉到,
它们参考了现行概率论中的一些最基本的概念。这些概念在指导我完成关于分段回归分
析的方法论构建时起了巨大的作用。正因为如此,我才说kolmogorov实在是太伟大了,
因为他在人类统计科学尚处于非常早期的时候就抽象出了那些概念。但是,作为一个非
数学背景的人士,要想透彻地理解那些已经完全脱离了实际物象的纯粹抽象的概念,我
实在是感到有些困难。于是才试图以自己独立的思考来提出一套简洁直观的概念系统。
我在版上开这一腔的目的就是想从大家的评论中重新思考是否恰当。我一再声明过,这
只是一次尝试,并非意味着我要和大家对抗。
is
theory
【在 s*****r 的大作中提到】 : Your "theory" is about random variable, right? defining a random variable is : not just saying the variable takes different values. If you are serious, : tell us how to define a random variable, either you know from classic : statistics, or your new theory. : if you don't understand the definition of rv, how could you develop a theory : about it? : : must learn math though it does discuss Statistics in it own style? I would : like to try another style. Why not? Did I really make mistakes from the : angle of mathematics? Did I really make mistakes from the angle of
|
s*****r 发帖数: 790 | 27 顾左右而言他。
不知所言。
你就直接回答问题不就得了?想想你当时追着要人回答问题。这么简单直接的没有任何
歧义的问题,还不能回答么?
【在 T*******I 的大作中提到】 : 我想我并非只是关心如何定义随机变量,而是定义那些我认为在统计学中最基础的几个 : 概念。这些概念是从人类最一般的统计行为中抽象出来的。这些行为包括了一般的抽样 : 调查和统计量的构造等。这些活动不是数学要关心的问题,因为数学不是一门研究统计 : 行为模式的学科,而是一门关于与“数学”有关的抽象概念的逻辑推理的学科。 : 我在这里组织的这些概念即使在非统计的人来看也是可以理解的。你们也可以感觉到, : 它们参考了现行概率论中的一些最基本的概念。这些概念在指导我完成关于分段回归分 : 析的方法论构建时起了巨大的作用。正因为如此,我才说kolmogorov实在是太伟大了, : 因为他在人类统计科学尚处于非常早期的时候就抽象出了那些概念。但是,作为一个非 : 数学背景的人士,要想透彻地理解那些已经完全脱离了实际物象的纯粹抽象的概念,我 : 实在是感到有些困难。于是才试图以自己独立的思考来提出一套简洁直观的概念系统。
|
y**t 发帖数: 205 | 28 Well, if you have questions in your research and want to ask on this forum,
I suggest you to rephrase the question in mathematical language. Just to
make sure people understand your question.
If you ask question just to test me, sorry, I don't do TA now.
one
a
【在 T*******I 的大作中提到】 : If what you said is true, you did a really great job. I wish I could be one : of the biological students. I must appreciate your job. However, I have : never had such a chance. : So, let me ask you, what is a difference between a measure in Math and in a : real activity of measurement? suppose I was one of your student. : : in
|
T*******I 发帖数: 5138 | 29 伟大的哲学家和数学家迪卡尔说过,不要相信任何既定的理论,放下你所掌握的从他人那里学到的全部知识,仅从一个或几个最简单直观的概念开始进行独立思考,你就会获得自己的认识。我想,我可能是无意识地这么做了。
我在作分段回归分析时也是如此,完全是独立思考的结果,特别是早期在国内思考这个问题时几乎没有受到任何现行方法的影响,否则,我很有可能在刚开始踏入这个领域时就会采纳强制连续性假设进而解方程求解临界点!如果真的如此,我可能什么也作不了了,因为那些数学方法天衣无缝,无懈可击。而每个人都把数学看得神圣无比,因为它被认为是科学的王冠。
当然,你们依然不会认同我的非解方程式求解临界点的办法。但它确实是真正的统计估计办法:有可测空间、有随机变异性、有分布、有期望并有可信区间,且所有这些都有可测性。而令人啼笑皆非的是,解方程式求解法却存在不可测性!且其变异性更不可测。我想正是因为如此,人们才引入了penalty的概念吧?(说实在的,不太懂这个penalty是什么东西,我是从自己的角度估摸的,因为我知道,在随机模拟试验下,解方程式求解临界点会发生极大的、不可测的变异性。)
,
【在 y**t 的大作中提到】 : Well, if you have questions in your research and want to ask on this forum, : I suggest you to rephrase the question in mathematical language. Just to : make sure people understand your question. : If you ask question just to test me, sorry, I don't do TA now. : : one : a
|
T*******I 发帖数: 5138 | 30 As you said, it is a set. However, a kolmogorov's sample space is also a set. Why can it be called a space? Why cannot the set that I defined be called a space? If you say that there is no repeat elements in the kolmogorov's sample space, either in my space there is no repeat individuals. Am I right?
,
【在 s*****r 的大作中提到】 : don't call it a space, it is just a set. : if you don't want people to think those terms in math, use your know terms, : like Actuaries suggested. : please don't use any mathematical terms, since you don't understand them. : you don't need to learn since you won't be able to understand anyway. : so just do NOT use them. : : I : the
|
|
|
s*****r 发帖数: 790 | 31 so you don't know what a space is.
http://en.wikipedia.org/wiki/Space_%28mathematics%29
In mathematics, a space is a set with some added structure.
did you see the "added structure"? what is your "added structure"?
set. Why can it be called a space? Why cannot the set that I defined be
called a space? If you say that there is no repeat elements in the
kolmogorov's sample space, either in my space there is no repeat individuals
. Am I right?
【在 T*******I 的大作中提到】 : As you said, it is a set. However, a kolmogorov's sample space is also a set. Why can it be called a space? Why cannot the set that I defined be called a space? If you say that there is no repeat elements in the kolmogorov's sample space, either in my space there is no repeat individuals. Am I right? : : ,
|
T*******I 发帖数: 5138 | 32 I said that all individuals in a population constitute a space called
population space. Ths space have included all elements and structures, nothing
more can be added into it.
【在 s*****r 的大作中提到】 : so you don't know what a space is. : http://en.wikipedia.org/wiki/Space_%28mathematics%29 : In mathematics, a space is a set with some added structure. : : did you see the "added structure"? what is your "added structure"? : : set. Why can it be called a space? Why cannot the set that I defined be : called a space? If you say that there is no repeat elements in the : kolmogorov's sample space, either in my space there is no repeat individuals : . Am I right?
|
d******e 发帖数: 7844 | 33 你真执著,跟大师理论要劳逸结合,调戏他一段时间,正经一段时间,这样才有意思
individuals
【在 s*****r 的大作中提到】 : so you don't know what a space is. : http://en.wikipedia.org/wiki/Space_%28mathematics%29 : In mathematics, a space is a set with some added structure. : : did you see the "added structure"? what is your "added structure"? : : set. Why can it be called a space? Why cannot the set that I defined be : called a space? If you say that there is no repeat elements in the : kolmogorov's sample space, either in my space there is no repeat individuals : . Am I right?
|
l********w 发帖数: 253 | 34
nothing
实在看不下去了,按照你的说法,我们为什么需要定义“space",不就用“set”就行
了?
【在 T*******I 的大作中提到】 : I said that all individuals in a population constitute a space called : population space. Ths space have included all elements and structures, nothing : more can be added into it.
|
T*******I 发帖数: 5138 | 35 Sorry, I can find nothing wrong in the definition about the "Populaiotn
space".
You said "实在看不下去了". This is your problem but not mine. I didn't think
that I make any mistake in logics.
【在 l********w 的大作中提到】 : : nothing : 实在看不下去了,按照你的说法,我们为什么需要定义“space",不就用“set”就行 : 了?
|
A*******s 发帖数: 3942 | 36 "there is nothing right in your left brain, and there is nothing left in
your right brain."
i can't help joking about u... sorry i am just toooooo bored...
think
【在 T*******I 的大作中提到】 : Sorry, I can find nothing wrong in the definition about the "Populaiotn : space". : You said "实在看不下去了". This is your problem but not mine. I didn't think : that I make any mistake in logics.
|
T*******I 发帖数: 5138 | 37 Nothing but joking to youself.
【在 A*******s 的大作中提到】 : "there is nothing right in your left brain, and there is nothing left in : your right brain." : i can't help joking about u... sorry i am just toooooo bored... : : think
|
T*******I 发帖数: 5138 | |
j*x 发帖数: 931 | 39 "So, let me ask you, what is a difference between a measure in Math and in a
real activity of measurement? suppose I was one of your student."
Let me try to answer this question.
Suppose we do "a real activity of measurement", say, the length of [0,1],
that is quite easy, 1. For [0,1/4] union with [3/4, 1], the length is 1/2.
Now, what if the set that we are "measuring" gets more complicated, such
that it is not simply the union of intervals? How are we going to "measure"
them? So we really need to define what "measure" is in a strict manner, so
that there will be no (or least) confusion.
In fact, no matter how you define measure, as long as it satisfies some "
must have" properties, there are always sets that can not be measured.
So, "real activity of measurement" is not always "actable":). |
T*******I 发帖数: 5138 | 40 我知道你是在从纯数学的高度抽象的角度来定义测度。我也不认为这个定义有何错误。
1997年11月至1998年3月底期间我曾在武汉大学的数学系旁听过几次测度论的课。听得
云里雾里,但对连续测度的数学定义还有一点点印象,记得那位老师口中常常出现一个
“左开右闭区间”术语,好像还有一连串这样的区间就可以构成一个连续测度等等。都
忘了。
我在这里提出的一个概念系统是一个非常简单、直观、朴素且经验主义化的,也就是说
它可以作为数学中严格定义相关的高级概念的初级概念。适用于中学和大学中一般非数
学专业的统计学教育。
a
"
【在 j*x 的大作中提到】 : "So, let me ask you, what is a difference between a measure in Math and in a : real activity of measurement? suppose I was one of your student." : Let me try to answer this question. : Suppose we do "a real activity of measurement", say, the length of [0,1], : that is quite easy, 1. For [0,1/4] union with [3/4, 1], the length is 1/2. : Now, what if the set that we are "measuring" gets more complicated, such : that it is not simply the union of intervals? How are we going to "measure" : them? So we really need to define what "measure" is in a strict manner, so : that there will be no (or least) confusion. : In fact, no matter how you define measure, as long as it satisfies some "
|
|
|
j*x 发帖数: 931 | 41 嗯,“非常简单、直观、朴素且经验主义化的”“概念系统”的确很好,但是很多时候
它们也只能解决“非常简单、直观、朴素且经验主义化的”问题。 |
T*******I 发帖数: 5138 | 42 是的,它们确实不太严谨。所以我说它们仅仅是初级概念。
【在 j*x 的大作中提到】 : 嗯,“非常简单、直观、朴素且经验主义化的”“概念系统”的确很好,但是很多时候 : 它们也只能解决“非常简单、直观、朴素且经验主义化的”问题。
|
m******2 发帖数: 564 | |