File size: 6,602 Bytes
b764195
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
\begin{thebibliography}{26}
\providecommand{\natexlab}[1]{#1}
\providecommand{\url}[1]{\texttt{#1}}
\expandafter\ifx\csname urlstyle\endcsname\relax
  \providecommand{\doi}[1]{doi: #1}\else
  \providecommand{\doi}{doi: \begingroup \urlstyle{rm}\Url}\fi

\bibitem[Agrawal(2023)]{agrawal2022bandits}
Agrawal, S.
\newblock \emph{Bandits with Heavy Tails: Algorithms Analysis and Optimality}.
\newblock PhD thesis, Tata Institute of Fundamental Research, 2023.
\newblock URL \url{http://hdl.handle.net/10603/478863}.

\bibitem[Agrawal \& Ramdas(2025)Agrawal and Ramdas]{agrawal2025stopping}
Agrawal, S. and Ramdas, A.
\newblock On stopping times of power-one sequential tests: Tight lower and upper bounds.
\newblock \emph{arXiv preprint arXiv:2504.19952}, 2025.

\bibitem[Agrawal et~al.(2020)Agrawal, Juneja, and Glynn]{agrawal2020optimal}
Agrawal, S., Juneja, S., and Glynn, P.
\newblock Optimal $\delta$-correct best-arm selection for heavy-tailed distributions.
\newblock In \emph{Algorithmic Learning Theory}, pp.\  61--110. PMLR, 2020.

\bibitem[Agrawal et~al.(2021{\natexlab{a}})Agrawal, Juneja, and Koolen]{pmlr-v134-agrawal21a}
Agrawal, S., Juneja, S.~K., and Koolen, W.~M.
\newblock Regret minimization in heavy-tailed bandits.
\newblock In \emph{Proceedings of Thirty Fourth Conference on Learning Theory}, volume 134 of \emph{Proceedings of Machine Learning Research}, pp.\  26--62. PMLR, 15--19 Aug 2021{\natexlab{a}}.

\bibitem[Agrawal et~al.(2021{\natexlab{b}})Agrawal, Koolen, and Juneja]{agrawal2021optimal}
Agrawal, S., Koolen, W.~M., and Juneja, S.
\newblock Optimal best-arm identification methods for tail-risk measures.
\newblock \emph{Advances in Neural Information Processing Systems}, 34:\penalty0 25578--25590, 2021{\natexlab{b}}.

\bibitem[Anscombe(1952)]{anscombe1952large}
Anscombe, F.~J.
\newblock Large-sample theory of sequential estimation.
\newblock In \emph{Mathematical Proceedings of the Cambridge Philosophical Society}, volume~48, pp.\  600--607. Cambridge University Press, 1952.

\bibitem[Asmussen(2003)]{asmussen2003applied}
Asmussen, S.
\newblock \emph{Applied probability and queues}.
\newblock Springer, 2003.

\bibitem[Billingsley(2017)]{billingsley2017probability}
Billingsley, P.
\newblock \emph{Probability and measure}.
\newblock John Wiley \& Sons, 2017.

\bibitem[Burnetas \& Katehakis(1996)Burnetas and Katehakis]{burnetas1996optimal}
Burnetas, A.~N. and Katehakis, M.~N.
\newblock Optimal adaptive policies for sequential allocation problems.
\newblock \emph{Advances in Applied Mathematics}, 17\penalty0 (2):\penalty0 122--142, 1996.

\bibitem[Chernoff(1959)]{chernoff1992sequential}
Chernoff, H.
\newblock Sequential design of experiments.
\newblock \emph{Ann. Math. Statist.}, 30\penalty0 (4):\penalty0 755--770, 1959.

\bibitem[Darling \& Robbins(1967)Darling and Robbins]{darling1967iterated}
Darling, D.~A. and Robbins, H.
\newblock Iterated logarithm inequalities.
\newblock \emph{Proceedings of the National Academy of Sciences}, 57\penalty0 (5):\penalty0 1188--1192, 1967.

\bibitem[Deep et~al.(2024)Deep, Bassamboo, and Juneja]{deep2024asymptotically}
Deep, V., Bassamboo, A., and Juneja, S.~K.
\newblock Asymptotically optimal and computationally efficient average treatment effect estimation in a/b testing.
\newblock 2024.

\bibitem[Deep et~al.(2025)Deep, Bassamboo, and Juneja]{deep2025asymptotic}
Deep, V., Bassamboo, A., and Juneja, S.
\newblock Asymptotic optimality theory of confidence intervals of the mean.
\newblock \emph{arXiv preprint arXiv:2501.19126}, 2025.

\bibitem[Fan \& Glynn(2025)Fan and Glynn]{fan2025fragility}
Fan, L. and Glynn, P.~W.
\newblock The fragility of optimized bandit algorithms.
\newblock \emph{Operations Research}, 73\penalty0 (6):\penalty0 3173--3198, 2025.

\bibitem[Gut(2009)]{gut2009stopped}
Gut, A.
\newblock \emph{Stopped random walks}.
\newblock Springer, 2009.

\bibitem[Honda \& Takemura(2010)Honda and Takemura]{honda2010asymptotically}
Honda, J. and Takemura, A.
\newblock An asymptotically optimal bandit algorithm for bounded support models.
\newblock In \emph{COLT}, pp.\  67--79, 2010.

\bibitem[Honda \& Takemura(2015)Honda and Takemura]{honda2015non}
Honda, J. and Takemura, A.
\newblock Non-asymptotic analysis of a new bandit algorithm for semi-bounded rewards.
\newblock \emph{J. Mach. Learn. Res.}, 16:\penalty0 3721--3756, 2015.

\bibitem[Jourdan et~al.(2022)Jourdan, Degenne, Baudry, de~Heide, and Kaufmann]{jourdan2022top}
Jourdan, M., Degenne, R., Baudry, D., de~Heide, R., and Kaufmann, E.
\newblock Top two algorithms revisited.
\newblock \emph{Advances in Neural Information Processing Systems}, 35:\penalty0 26791--26803, 2022.

\bibitem[Lai \& Robbins(1985)Lai and Robbins]{lai1985asymptotically}
Lai, T.~L. and Robbins, H.
\newblock Asymptotically efficient adaptive allocation rules.
\newblock \emph{Advances in Applied Mathematics}, 6\penalty0 (1):\penalty0 4--22, 1985.

\bibitem[Mukhopadhyay(2020)]{mukhopadhyay2020asymptotic}
Mukhopadhyay, N.
\newblock Asymptotic normality of sequential stopping times with applications: Confidence intervals for an exponential mean.
\newblock \emph{Calcutta Statistical Association Bulletin}, 72\penalty0 (1):\penalty0 17--34, 2020.

\bibitem[Panda \& Agrawal(2026)Panda and Agrawal]{panda2026regret}
Panda, S. and Agrawal, S.
\newblock Regret tail characterization of optimal bandit algorithms with generic rewards.
\newblock \emph{arXiv preprint arXiv:2604.14876}, 2026.

\bibitem[Robbins \& Siegmund(1974)Robbins and Siegmund]{robbins1974expected}
Robbins, H. and Siegmund, D.
\newblock The expected sample size of some tests of power one.
\newblock \emph{The Annals of Statistics}, 2\penalty0 (3):\penalty0 415--436, 1974.

\bibitem[Siegmund(2013)]{siegmund2013sequential}
Siegmund, D.
\newblock \emph{Sequential Analysis: Tests and Confidence Intervals}.
\newblock Springer Science \& Business Media, 2013.

\bibitem[Wald(1992)]{wald1992sequential}
Wald, A.
\newblock Sequential tests of statistical hypotheses.
\newblock In \emph{Breakthroughs in statistics: Foundations and basic theory}, pp.\  256--298. Springer, 1992.

\bibitem[Wald \& Wolfowitz(1948)Wald and Wolfowitz]{wald1948optimum}
Wald, A. and Wolfowitz, J.
\newblock Optimum character of the sequential probability ratio test.
\newblock \emph{The Annals of Mathematical Statistics}, pp.\  326--339, 1948.

\bibitem[Wang et~al.(2026)Wang, Agrawal, and Ramdas]{wang2026almost}
Wang, H., Agrawal, S., and Ramdas, A.
\newblock Almost sure null bankruptcy of testing-by-betting strategies.
\newblock \emph{arXiv preprint arXiv:2602.08888}, 2026.

\end{thebibliography}