User:Gesalbte
Our first steps tour and our frequently asked questions will help you a lot after registration. They explain how to customize the interface (for example the language), how to upload files and our basic licensing policy (Wikimedia Commons only accepts free content). You don't need technical skills in order to contribute here. Be bold when contributing and assume good faith when interacting with others. This is a wiki. More information is available at the community portal. You may ask questions at the help desk, village pump or on IRC channel #wikimedia-commons (webchat). You can also contact an administrator on their talk page. If you have a specific copyright question, ask at the copyright village pump. |
|
Introduction
[edit]Born into Uncyclopedia, this user is an extraordinary dumbfuck[1], who has messed up everybody in Spring Weekend[2].
This user is currently a student studying finance in college and enjoys the sunshine, rain, and forest of Connecticut. If you have to excuse him, please leave message in the discussion page.
This user is a dumbfuck writer[3] and wrote a lot of stuff that appreciated by his professor[4]. For example, the sentence below comes from Nostra Via in Regnum Ducit:
“ | Only after thou hast suffered the darkened night, there cometh the light. | ” |
Statistics
[edit]User:Disavian/Userboxes/Ubx-Night
Chapter 2
[edit]inter-quartile range (IQR; 四分位距); Outlier: |z|>2; ext: |z|>3
Chebychev’s Theorem: at least data is within ( ), Empirical Rule: approx. 68% at z ∈ (-1s, 1s), approx. 95% at z ∈ (-2s, 2s), approx. 99.7% at z ∈ (-3s, 3s)
Chapter 3
[edit]: from n choose (组合) k; : from n permute (排列) k.
- intersection, : union, Ac: complementary. ; factorials (n!);
Chapter 4
[edit]- Discrete random variables (rv); STAT → CALC → 1-Var Stats → L1(值), L2(个数)
Binomial r.v. (二项分布; Bernoulli trial: coin-toss); =math. expectation=np,
Probability Density Function: ; DISTR → binompdf (n, p, k). [n个成功率为p的伯努利实验,成功k次的概率是多少?] Cumulative Distribution Function: DISTR → binomcdf (n, p, k) [n个成功率为p的伯努利实验,至少成功k次的概率是多少?]
Uniform (平均) distribution: ; ; ; .
Normal distribution (正态分布): Z ~ N (), ~ N (0, 1); std. norm.: N (0, 1);
Cumulative Distribution Function: DISTR → normalcdf (a, b, μ, σ). [对于正态分布N (),作下限为a,上限为b的定积分得多少?] Quantile function: DISTR → invnorm (p, μ, σ). [对于正态分布N (),作积分下限为-∞的定积分得p时,积分上限是多少?]
Criteria for determining whether normal distribution: 1) histogram or stem and leaf display is bell shaped; 2) data satisfies empirical rule; 3) IQR/s = (Q3-Q1)/s is approximately 1.3; 4) a normal probability plot is approximately linear. [TI: 1) clear y functions 2) enter data 3) 2ND → Y= → ... → 6th plot, x axis 4) ZOOM 9]
Central Limit Theorem: 许多(n>30)平均值标准差的分布,其平均值的分布为N (, ).
Chapter 5: Jargon Box: confidence coefficient(置信系数; 1-α), confidence level(置信度; 100×(1-α)).
A. 根据样本猜总体的平均值μ: = , where = invT(1-/2, n-1), n-1是自由度.
Also that T-distribution function (normal + centered at 0, fatter tails; df = n-1↑tend to be normal). Std. dev. of T-distribution is , its Cumulative Distribution Function: DISTR → tcdf(a, b, df), where df=n-1 [对于自由度df的T分布,作下限为a,上限为b的定积分得多少?]; its Quantile function: DISTR → invT(p, df) [对于自由度为df的T分布,作积分下限为-∞的定积分得p时,积分上限是多少?].
近似: When it’s large (n≥30) and σ is known, we can use z-CI instead: = ,
where = invNorm(1-/2), 1-/2 = (1-CC)/2. [ = 1.645; = 1.960; = 2.576]
已知置信区间,求最小样本容量: for z-CI, , therefore . Why not t-CI ( )? Because n-1 in you cannot bring it out.
B. 根据样本的比例猜总体的比例p: p , when we have large samples ( , ), there , . This function is based on ; , where .
When p nears 0 or 1, use adjusted confidence interval , where . Sampling error (SE; SE=.5Width) or margin of error (ME): , therefore .
Chapter 6
[edit]Hypothesis testing [无论如何,总体分布必须为正态的时候才能检定]
Type I error: rejected a correct H0; Type II error: failed to reject a wrong H0. The smaller
selected, the more evidence (larger z) needed to reject H0. 思想罪: 思想就是犯罪![5]
A. t-test: H0说μ0, 你不相信, 就搞了Ha: , s, n, 代入下面这个公式, 看看你的图像牛逼不?
- where is the sample mean, μ0 is the claimed mean (=H0), s is sample std.dev.
Left tailed test: Ha: μ < μ0; reject H0 if t < -t = -invT(1-α, n-1); p-value = tcdf(-10^99, t, n-1). Two tailed test: Ha: μ ≠ μ0; reject H0 if t [-t, t], where t = invT(1-α/2, n-1); p-value = 2×tcdf(|t|, 10^99, n-1). Right tailed test: Ha: μ > μ0; reject H0 if t > t = invT(1-α, n-1); p-value = tcdf(t, 10^99, n-1).
近似: z-test: when σ is known and sample size is very large (n>30).
- where is the sample mean, μ0 is the claimed mean (=H0), σ is population std.dev.
Left tailed test: Ha: μ < μ0; reject H0 if z < -z = -invNorm(1-α); p-value = normalcdf(-10^99, z). Two tailed test: Ha: μ ≠ μ0; reject H0 if z [-z, z], where z = invNorm(1-α/2); p-value = 2×normalcdf(|z|, 10^99). Right tailed test: Ha: μ > μ0; reject H0 if z > z = invNorm(1-α); p-value = normalcdf(z, 10^99).
B. z-test for population proportion: ; .
- where is sample proportion, p0 is the claimed mean proportion (=H0).
Left tailed test: Ha: p < p 0; reject H0 if z < -z = -invNorm(1-α); p-value = normalcdf(-10^99, z). Two tailed test: Ha: p ≠ p0; reject H0 if z [-z, z], where z = invNorm(1-α/2); p-value = 2×normalcdf(|z|, 10^99). Right tailed test: Ha: p > p 0; reject H0 if z > z = invNorm(1-α); p-value = normalcdf(z, 10^99).
Chapter 7
[edit]Large sample: CI = ; Test statistic:
One tailed test:
H0: (μ1-μ2) = D0
Ha: (μ1-μ2) < D0
[or Ha: (μ1-μ2) > D0]
Two tailed test:
H0: (μ1-μ2) = D0
Ha: (μ1-μ2) ≠ D0
where D0 = hypothesized difference between the means (often it is equal to 0)
Rejection region: z < -
[or z > ]
Rejection region: |z| >
t-test: CI = , where , = invT(1-/2, )
- proportion test
- CI = ,
, CI = ; , CI = ; where , 其他就是差.
Test
[edit]The person is out of his mind. User:Fastily/Userboxes/Hopelesseditingaddict User:Strdst grl/ubx/mandelbrot User:Teinesavaii/Polynesian Userboxes/Shrink User:UBX/For rent User:Strdst grl/ubx/bouncing User:Edit Centric/UBX/ArmBears User:P.B. Pilhet/UBX/NDC
Major
[edit]Minor
[edit]Test Reference
[edit]This section is where I'm testing the citation codes, to see if they are correctly written.
Hooray!
[edit]- ↑ Philip talked to Angela: Oh my God! This chink is an an extraordinary dumbfuck!
- ↑ Andrew: Yeah! Get some pussies. You're the man, I appreciate that.
- ↑ Philip: What the f- is wrong with you? What the f- is this? This is English? Are you f-ing retard?
- ↑ Who told you that?
- ↑ G. Orwell: 1984. Crimestop your thinkcrime! You just committed a facecrime!