Skip to main content
edited answere to question #3
Source Link
Chee
  • 984
  • 5
  • 18

Update (April 21, 2017): answer to question #3 The convergence rate of uniform order statistics to their expecations obtained from $(1)$ is NOT optimal for order statisic $X_{(k)}$ for which $k=o(n)$$k$ or $k/n \to 1$$n-k$ is bounded. However, it is optimal for $X_{(k)}$ for which $k= c_n n$ with $$1 > \limsup c_n \ge \liminf c_n >0.$$ In other words, close to the edges 0 or 1, the optimal reate of convergence is proportional to $$\frac{\log n}{n},$$ whereas in other parts of the compact interval $[0,1]$ the optimal rate of convergence is $$\frac{\sqrt{2 \log_{(2)}n}}{\sqrt{n}}.$$ This can be seen from:

Update (April 21, 2017): answer to question #3 The convergence rate of uniform order statistics to their expecations obtained from $(1)$ is NOT optimal for order statisic $X_{(k)}$ for which $k=o(n)$ or $k/n \to 1$. However, it is optimal for $X_{(k)}$ for which $k= c_n n$ with $$1 > \limsup c_n \ge \liminf c_n >0.$$ In other words, close to the edges 0 or 1, the optimal reate of convergence is proportional to $$\frac{\log n}{n},$$ whereas in other parts of the compact interval $[0,1]$ the optimal rate of convergence is $$\frac{\sqrt{2 \log_{(2)}n}}{\sqrt{n}}.$$ This can be seen from:

Update (April 21, 2017): answer to question #3 The convergence rate of uniform order statistics to their expecations obtained from $(1)$ is NOT optimal for order statisic $X_{(k)}$ for which $k$ or $n-k$ is bounded. However, it is optimal for $X_{(k)}$ for which $k= c_n n$ with $$1 > \limsup c_n \ge \liminf c_n >0.$$ In other words, close to the edges 0 or 1, the optimal reate of convergence is proportional to $$\frac{\log n}{n},$$ whereas in other parts of the compact interval $[0,1]$ the optimal rate of convergence is $$\frac{\sqrt{2 \log_{(2)}n}}{\sqrt{n}}.$$ This can be seen from:

chagned title of post back; added answere to question #3
Source Link
Chee
  • 984
  • 5
  • 18

Locations Rate of convergence of uniform order statistics: two rates of convergence and one mystery to their expectations

Update to question #3 (April 21, 20162017): answer to question #3 The convergence rate of uniform order statistics to their expecations obtained from $(1)$ is NOT optimal for order statisic $X_{(k)}$ for which $k=o(n)$ or $k/n \to 1$. However, it is optimal for $X_{(k)}$ for which $k= c_n n$ with $$1 > \limsup c_n \ge \liminf c_n >0.$$ In other words, close to the edges 0 or 1, the optimal reate of convergence is proportional to $$\frac{\log n}{n},$$ whereas in other parts of the compact interval $[0,1]$ the optimal rate of convergence is $$\frac{\sqrt{2 \log_{(2)}n}}{\sqrt{n}}.$$ This can be seen from:

Lemma A2.1 on page 148 of "asymptotic expansions for the power of distribution free tests in the one-sample problem" by Albers, Bickel and Zwet.

Locations of uniform order statistics: two rates of convergence and one mystery

Update to question #3 (April 21, 2016)

Rate of convergence of uniform order statistics to their expectations

Update (April 21, 2017): answer to question #3 The convergence rate of uniform order statistics to their expecations obtained from $(1)$ is NOT optimal for order statisic $X_{(k)}$ for which $k=o(n)$ or $k/n \to 1$. However, it is optimal for $X_{(k)}$ for which $k= c_n n$ with $$1 > \limsup c_n \ge \liminf c_n >0.$$ In other words, close to the edges 0 or 1, the optimal reate of convergence is proportional to $$\frac{\log n}{n},$$ whereas in other parts of the compact interval $[0,1]$ the optimal rate of convergence is $$\frac{\sqrt{2 \log_{(2)}n}}{\sqrt{n}}.$$ This can be seen from:

Lemma A2.1 on page 148 of "asymptotic expansions for the power of distribution free tests in the one-sample problem" by Albers, Bickel and Zwet.

added 46 characters in body
Source Link
Chee
  • 984
  • 5
  • 18

Update to question #3 (April 21, 2016)

Update to question #3 (April 21, 2016)

edit deviation of uniform order statistics from their classic locations; removed a comment in question #3
Source Link
Chee
  • 984
  • 5
  • 18
Loading
highlighted maximal spacing and part of question #3
Source Link
Chee
  • 984
  • 5
  • 18
Loading
edited discussion in question #3
Source Link
Chee
  • 984
  • 5
  • 18
Loading
edited title; expanded discussion; raised question #3
Source Link
Chee
  • 984
  • 5
  • 18
Loading
restructured updated post; expanded post; edited question
Source Link
Chee
  • 984
  • 5
  • 18
Loading
added 21 characters in body
Source Link
Chee
  • 984
  • 5
  • 18
Loading
updated rate of convergence; added a question on maximal uniform spacing
Source Link
Chee
  • 984
  • 5
  • 18
Loading
edited tags
Link
Henry.L
  • 8.1k
  • 8
  • 48
  • 74
Loading
add second update
Source Link
Chee
  • 984
  • 5
  • 18
Loading
update on what the reference is about
Source Link
Chee
  • 984
  • 5
  • 18
Loading
changed equality to inequality wrt beta in question
Source Link
Chee
  • 984
  • 5
  • 18
Loading
added a reference
Source Link
Chee
  • 984
  • 5
  • 18
Loading
add condition on beta
Source Link
Chee
  • 984
  • 5
  • 18
Loading
added 10 characters in body
Source Link
Chee
  • 984
  • 5
  • 18
Loading
added comments to qustion
Source Link
Chee
  • 984
  • 5
  • 18
Loading
Source Link
Chee
  • 984
  • 5
  • 18
Loading