Maximum likelihood estimation of burst-merging kernels for bursty time series

Hang-Hyun Jo; Tibebe Birhanu

arxiv: 2503.14861 · v1 · submitted 2025-03-19 · ⚛️ physics.data-an

Maximum likelihood estimation of burst-merging kernels for bursty time series

Tibebe Birhanu , Hang-Hyun Jo This is my paper

Pith reviewed 2026-05-23 00:46 UTC · model grok-4.3

classification ⚛️ physics.data-an

keywords burst-merging kernelmaximum likelihood estimationbursty time seriesburst treehierarchical mergingempirical time seriestimescale analysis

0 comments

The pith

Maximum likelihood estimation recovers the burst-merging kernel from observed time series.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a maximum likelihood estimation method to recover the burst-merging kernel directly from time series data. This kernel encodes the rule by which individual events merge into small bursts and then larger ones as the defining timescale increases. A sympathetic reader would care because the method turns the full hierarchical burst tree into a single estimable object that can be compared across datasets. The approach is validated by recovering known kernels from synthetic series generated by several model kernels and is demonstrated on empirical series from varied domains. If successful, the method supplies a quantitative description that supports more precise investigation of the processes producing burstiness.

Core claim

The burst-tree structure can be simply characterized by a burst-merging kernel that dictates which bursts are merged together as the timescale increases. We develop the maximum likelihood estimation method of the burst-merging kernel from time series, which is successfully tested against the time series generated using several model kernels. We also apply our method to some empirical time series from various backgrounds.

What carries the argument

The burst-merging kernel, a function that determines which bursts merge as timescale coarsens, recovered by maximum likelihood estimation from the observed event times.

If this is right

The estimated kernel supplies a precise quantitative characterization of any given time series.
Comparison of kernels across datasets becomes possible once the estimation procedure is applied.
Underlying generative mechanisms can be studied more accurately once the kernel is known.
The method works for both synthetic series drawn from known kernels and real empirical series.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Kernels recovered from different domains could be compared to test whether bursty processes share common merging rules.
The estimation procedure could be extended to time series with slowly varying statistics by allowing the kernel to depend on external covariates.
If the kernel turns out to be low-dimensional in many systems, it would reduce the effective degrees of freedom needed to model burst hierarchies.

Load-bearing premise

The hierarchical merging pattern of bursts is fully captured by a single kernel depending on burst properties.

What would settle it

Forward simulation of new event series from the estimated kernel that produces burst trees whose merging statistics deviate from those measured in the original data.

Figures

Figures reproduced from arXiv: 2503.14861 by Hang-Hyun Jo, Tibebe Birhanu.

**Figure 2.** Figure 2: FIG. 2. Schematic diagram of the merging process for the [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗

**Figure 3.** Figure 3: FIG. 3. Model kernels in Eqs. (11)–(14) (left) used to gen [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

**Figure 4.** Figure 4: FIG. 4. Estimated kernels from four empirical time series [PITH_FULL_IMAGE:figures/full_fig_p005_4.png] view at source ↗

read the original abstract

Various time series in natural and social processes have been found to be bursty. Events in the time series rapidly occur within short time periods, forming bursts, which are alternated with long inactive periods. As the timescale defining bursts increases, individual events are sequentially merged to become small bursts and then bigger ones, eventually leading to the single burst containing all events. Such a merging pattern has been depicted by a tree that fully reveals the hierarchical structure of bursts, thus called a burst tree. The burst-tree structure can be simply characterized by a burst-merging kernel that dictates which bursts are merged together as the timescale increases. In this work, we develop the maximum likelihood estimation method of the burst-merging kernel from time series, which is successfully tested against the time series generated using several model kernels. We also apply our method to some empirical time series from various backgrounds. Our method provides a useful tool to precisely characterize the time series data, hence enabling to study their underlying mechanisms more accurately.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper gives an MLE for burst-merging kernels that recovers inputs on synthetic data but leaves the single-kernel assumption untested on real series.

read the letter

The main point is that the authors built a maximum likelihood estimator for the burst-merging kernel that describes how events merge into larger bursts as the observation timescale grows. They recover the correct kernel when the data are generated from the same model kernels they fit, and they show example fits on a few empirical series from different domains. That is the concrete advance over earlier burst-tree descriptions. The synthetic recovery checks that the optimization step works as intended once the tree is built from the event times. The empirical runs give a first sense of what the fitted kernels look like in practice. Both steps are useful for anyone who already accepts the burst-tree framing and wants a fitting procedure rather than hand-chosen kernels. The limitation is the one flagged in the stress-test note. Every recovery test uses data produced by the exact model class being estimated, so the exercise confirms numerical correctness under the modeling assumption but does not check whether real bursty series are generated by any single kernel. If merging rules depend on event size, carry memory, or require more than one functional form, the estimated kernel becomes a summary statistic rather than evidence about the process. The paper does not report any such checks or comparisons against richer alternatives. The citation list stays within the expected bursty-time-series literature and shows no obvious gaps or padding. The setup is not circular; the kernel is inferred from the observed times rather than defined in terms of itself. This paper is for specialists already working on hierarchical burst structure in physics or social data who need a quantitative fitting tool. A reader outside that niche will not find foundational claims or broad methodological lessons. It deserves peer review because the estimator is new, the synthetic tests are positive, and the empirical examples are a start; referees can ask for additional validation or robustness checks without the work being dismissed outright.

Referee Report

1 major / 2 minor

Summary. The manuscript develops a maximum likelihood estimation (MLE) method to recover the burst-merging kernel that governs the hierarchical merging of events into bursts as the observation timescale increases. The kernel is estimated directly from observed event times. The procedure is validated by generating synthetic time series from several chosen model kernels and recovering the input kernels; the method is then applied to empirical bursty time series from multiple domains.

Significance. If the estimator is correctly derived and implemented, the approach supplies a compact, data-driven characterization of burst-tree structure that goes beyond conventional inter-event time statistics. Successful synthetic recovery under the modeling assumptions demonstrates that the numerical procedure is feasible and can serve as a quantitative tool for comparing bursty processes across systems.

major comments (1)

[synthetic tests and empirical application] The validation consists exclusively of recovering kernels from data generated by the same family of model kernels used to define the likelihood (abstract and synthetic-test section). This confirms numerical correctness of the MLE under the modeling assumption but leaves untested whether any single kernel suffices to describe merging statistics in data generated by other processes or in empirical series; if the single-kernel assumption fails, the estimated kernel is not a sufficient statistic for the observed burst tree.

minor comments (2)

[abstract] The abstract states that the method is 'successfully tested' but supplies no explicit form of the likelihood, no optimization details, and no quantitative recovery metrics (e.g., parameter error or likelihood ratio); including these would allow immediate assessment of the central technical contribution.
[empirical results] When the estimated kernel is applied to empirical series, report at least the maximized log-likelihood value and a comparison against a baseline (e.g., uniform or exponential kernel) to quantify the improvement.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for the careful reading and constructive feedback on our manuscript. We respond to the major comment below.

read point-by-point responses

Referee: [synthetic tests and empirical application] The validation consists exclusively of recovering kernels from data generated by the same family of model kernels used to define the likelihood (abstract and synthetic-test section). This confirms numerical correctness of the MLE under the modeling assumption but leaves untested whether any single kernel suffices to describe merging statistics in data generated by other processes or in empirical series; if the single-kernel assumption fails, the estimated kernel is not a sufficient statistic for the observed burst tree.

Authors: We agree that the synthetic tests verify numerical correctness and unbiased recovery of the kernel when data are generated exactly from the assumed model family; this is the conventional validation for an MLE procedure. The empirical applications demonstrate practical use of the estimator on real bursty series, yielding a compact characterization under the model. The manuscript does not claim that a single kernel is always sufficient for arbitrary generative processes or that the estimate is invariably a sufficient statistic; it is presented as the maximum-likelihood description within the model class. When the assumption is violated, the fitted kernel remains the best approximation under the model (with possible diagnostics via likelihood or residuals), but we do not test model misspecification here. Broadening the validation to other generative mechanisms lies beyond the present scope. revision: no

Circularity Check

0 steps flagged

No circularity: MLE estimator derives kernel from observed event times without self-definition or fitted-input reduction

full rationale

The paper introduces a maximum likelihood procedure that takes raw event timestamps as input and outputs an estimated burst-merging kernel. Validation consists of forward simulation from chosen kernels followed by recovery; this tests numerical correctness of the estimator under the modeling assumption but does not make the target kernel a function of itself or rename fitted parameters as predictions. No self-citation chain, ansatz smuggling, or uniqueness theorem is invoked to close the derivation. The central claim therefore remains an independent statistical construction rather than a tautology.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

Review performed from abstract only; the ledger therefore records only the structural assumptions stated in the abstract.

axioms (1)

domain assumption The burst-tree structure can be simply characterized by a burst-merging kernel that dictates which bursts are merged together as the timescale increases.
Explicitly stated in the abstract as the modeling premise that enables the kernel estimation task.

pith-pipeline@v0.9.0 · 5699 in / 1180 out tokens · 40011 ms · 2026-05-23T00:46:46.030677+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

52 extracted references · 52 canonical work pages

[1]

At the initial time step s = 0, each event makes its own burst of size one

work page
[2]

At each time step s, two bursts of sizes b and b′ are chosen at random with a probability proportional to Kbb′ [Eq. (5)]

work page
[3]

Two merged bursts are assigned to left and right children nodes at random

Those two bursts are merged to make another burst of size b + b′, which becomes a parent node. Two merged bursts are assigned to left and right children nodes at random

work page
[4]

Once the ordinal burst tree is generated, we randomly draw n − 1 IETs from the power-law IET distribution as P (τ) = τ −α Pτc τ =1 τ −α for τ = 1,

Steps 2–3 are repeated until all events belong to the single burst of size n. Once the ordinal burst tree is generated, we randomly draw n − 1 IETs from the power-law IET distribution as P (τ) = τ −α Pτc τ =1 τ −α for τ = 1, . . . , τc, (15) where α is the power-law exponent and τc is the upper bound. These IETs are sorted in a descending order to obtain ...

work page
[5]

Barab´ asi and M

A.-L. Barab´ asi and M. P´ osfai,Network Science (Cam- bridge University Press, Cambridge, 2016)

work page 2016
[6]

M. E. J. Newman, Networks, second edition ed. (Oxford University Press, Oxford, United Kingdom; New York, NY, United States of America, 2018)

work page 2018
[7]

Menczer, S

F. Menczer, S. Fortunato, and C. A. Davis, A First Course in Network Science (Cambridge University Press, Cambridge, 2020)

work page 2020
[8]

S. N. Dorogovtsev and J. F. F. Mendes, The Nature of Complex Networks (Oxford University Press, Oxford, UK; New York, NY, 2022)

work page 2022
[9]

Easley and J

D. Easley and J. Kleinberg, Networks, Crowds, and Mar- kets: Reasoning About a Highly Connected World (Cam- bridge University Press, 2010)

work page 2010
[10]

Barrat, M

A. Barrat, M. Barth´ elemy, and A. Vespignani,Dynami- cal Processes on Complex Networks , 1st ed. (Cambridge University Press, 2008)

work page 2008
[11]

Holme and J

P. Holme and J. Saram¨ aki, Temporal networks, Physics Reports 519, 97 (2012)

work page 2012
[12]

Holme and J

P. Holme and J. Saram¨ aki, eds.,Temporal Network The- ory, Computational Social Sciences (Springer Interna- tional Publishing, Cham, 2019)

work page 2019
[13]

Masuda and R

N. Masuda and R. Lambiotte, A Guide to Temporal Net- works, Series on Complexity Science (World Scientific, New Jersey, 2016)

work page 2016
[14]

Barab´ asi, The origin of bursts and heavy tails in human dynamics, Nature 435, 207 (2005)

A.-L. Barab´ asi, The origin of bursts and heavy tails in human dynamics, Nature 435, 207 (2005)

work page 2005
[15]

Karsai, H.-H

M. Karsai, H.-H. Jo, and K. Kaski, Bursty Human Dy- namics (Springer International Publishing, Cham, 2018)

work page 2018
[16]

Jo and T

H.-H. Jo and T. Hiraoka, Bursty Time Series Analysis for Temporal Networks, in Temporal Network Theory , edited by P. Holme and J. Saram¨ aki (Springer Interna- tional Publishing, Cham, 2023) 2nd ed., pp. 165–183

work page 2023
[17]

P. Bak, C. Tang, and K. Wiesenfeld, Self-organized crit- icality: An explanation of the 1/f noise, Physical Review Letters 59, 381 (1987)

work page 1987
[18]

H. J. Jensen, Self-Organized Criticality: Emergent Com- plex Behavior in Physical and Biological Systems , 1st ed. (Cambridge University Press, 1998)

work page 1998
[19]

M. S. Wheatland, P. A. Sturrock, and J. M. McTiernan, The Waiting-Time Distribution of Solar Flare Hard X- Ray Bursts, The Astrophysical Journal 509, 448 (1998)

work page 1998
[20]

Corral, Long-term clustering, scaling, and universal- ity in the temporal occurrence of earthquakes, Physical Review Letters 92, 108501 (2004)

´A. Corral, Long-term clustering, scaling, and universal- ity in the temporal occurrence of earthquakes, Physical Review Letters 92, 108501 (2004)

work page 2004
[21]

J. M. Beggs and D. Plenz, Neuronal Avalanches in Neo- cortical Circuits, The Journal of Neuroscience 23, 11167 (2003)

work page 2003
[22]

Petermann, T

T. Petermann, T. C. Thiagarajan, M. A. Lebedev, M. A. L. Nicolelis, D. R. Chialvo, and D. Plenz, Spon- taneous cortical activity in awake monkeys composed of neuronal avalanches, Proceedings of the National Academy of Sciences 106, 15921 (2009)

work page 2009
[23]

Kemuriyama, H

T. Kemuriyama, H. Ohta, Y. Sato, S. Maruyama, M. Tandai-Hiruma, K. Kato, and Y. Nishida, A power- law distribution of inter-spike intervals in renal sym- pathetic nerve activity in salt-sensitive hypertension- induced chronic heart failure, BioSystems 101, 144 (2010)

work page 2010
[24]

Goh and A.-L

K.-I. Goh and A.-L. Barab´ asi, Burstiness and memory in complex systems, EPL (Europhysics Letters) 81, 48002 (2008)

work page 2008
[25]

Crane and D

R. Crane and D. Sornette, Robust dynamic classes re- vealed by measuring the response function of a social system, Proceedings of the National Academy of Sciences 105, 15649 (2008)

work page 2008
[26]

Rybski, S

D. Rybski, S. V. Buldyrev, S. Havlin, F. Liljeros, and H. A. Makse, Scaling laws of human interaction activity, Proceedings of the National Academy of Sciences 106, 12640 (2009)

work page 2009
[27]

H.-H. Jo, T. Hiraoka, and M. Kivel¨ a, Burst-tree decom- position of time series reveals the structure of temporal correlations, Scientific Reports 10, 12202 (2020)

work page 2020
[28]

J. Choi, T. Hiraoka, and H.-H. Jo, Individual-driven versus interaction-driven burstiness in human dynamics: The case of Wikipedia edit history, Physical Review E 104, 014312 (2021). 8

work page 2021
[29]

Karsai, K

M. Karsai, K. Kaski, A.-L. Barab´ asi, and J. Kert´ esz, Uni- versal features of correlated bursty behaviour, Scientific Reports 2, 397 (2012)

work page 2012
[30]

Birhanu and H.-H

T. Birhanu and H.-H. Jo, Burst-tree structure and higher-order temporal correlations, Physical Review E 111, 014308 (2025)

work page 2025
[31]

T. Pham, P. Sheridan, and H. Shimodaira, PAFit: A Sta- tistical Method for Measuring Preferential Attachment in Temporal Complex Networks, PLoS ONE 10, e0137796 (2015)

work page 2015
[32]

Barab´ asi and R

A.-L. Barab´ asi and R. Albert, Emergence of Scaling in Random Networks, Science 286, 509 (1999)

work page 1999
[33]

P. L. Krapivsky, S. Redner, and F. Leyvraz, Connectivity of Growing Random Networks, Physical Review Letters 85, 4629 (2000)

work page 2000
[34]

M. E. J. Newman, Clustering and preferential attach- ment in growing networks, Physical Review E64, 025102 (2001)

work page 2001
[35]

Jeong, Z

H. Jeong, Z. N´ eda, and A. L. Barab´ asi, Measuring prefer- ential attachment in evolving networks, Europhysics Let- ters (EPL) 61, 567 (2003)

work page 2003
[36]

Sheridan, Y

P. Sheridan, Y. Yagahara, and H. Shimodaira, Measur- ing preferential attachment in growing networks with missing-timelines using Markov chain Monte Carlo, Phys- ica A: Statistical Mechanics and its Applications 391, 5031 (2012)

work page 2012
[37]

Kunegis, M

J. Kunegis, M. Blattner, and C. Moser, Preferential at- tachment in online networks: Measurement and explana- tions, in Proceedings of the 5th Annual ACM Web Science Conference (ACM, Paris France, 2013) pp. 205–214

work page 2013
[38]

W. H. Stockmayer, Theory of Molecular Size Distribu- tion and Gel Formation in Branched-Chain Polymers, The Journal of Chemical Physics 11, 45 (1943)

work page 1943
[39]

Lushnikov, Evolution of coagulating systems, Journal of Colloid and Interface Science 45, 549 (1973)

A. Lushnikov, Evolution of coagulating systems, Journal of Colloid and Interface Science 45, 549 (1973)

work page 1973
[40]

W. H. White, On the form of steady-state solutions to the coagulation equations, Journal of Colloid and Interface Science 87, 204 (1982)

work page 1982
[41]

E. M. Hendriks, M. H. Ernst, and R. M. Ziff, Coagulation equations with gelation, Journal of Statistical Physics31, 519 (1983)

work page 1983
[42]

D. J. Aldous, Deterministic and Stochastic Models for Coalescence (Aggregation and Coagulation): A Review of the Mean-Field Theory for Probabilists, Bernoulli 5, 3 (1999)

work page 1999
[43]

M. H. Lee, A survey of numerical solutions to the coagu- lation equation, Journal of Physics A: Mathematical and General 34, 10219 (2001)

work page 2001
[44]

Leyvraz, Scaling theory and exactly solved models in the kinetics of irreversible aggregation, Physics Reports 383, 95 (2003)

F. Leyvraz, Scaling theory and exactly solved models in the kinetics of irreversible aggregation, Physics Reports 383, 95 (2003)

work page 2003
[45]

Leyvraz, Rigorous Results in the Scaling Theory of Irreversible Aggregation Kinetics:, Journal of Nonlinear Mathematical Physics 12, 449 (2005)

F. Leyvraz, Rigorous Results in the Scaling Theory of Irreversible Aggregation Kinetics:, Journal of Nonlinear Mathematical Physics 12, 449 (2005)

work page 2005
[46]

J. A. Wattis, An introduction to mathematical models of coagulation–fragmentation processes: A discrete deter- ministic mean-field approach, Physica D: Nonlinear Phe- nomena 222, 1 (2006)

work page 2006
[47]

Birhanu and H.-H

T. Birhanu and H.-H. Jo, Codes for the maximum likelihood estimation method of burst-merging kernels, https://github.com/tibebe22/Kernel-estimate (2025)

work page 2025
[48]

English Wikipedia, https://dumps.wikimedia.org/

work page
[49]

Yang and J

J. Yang and J. Leskovec, Patterns of temporal variation in online media, in Proceedings of the Fourth ACM In- ternational Conference on Web Search and Data Mining (ACM Press, Hong Kong, China, 2011) pp. 177–186

work page 2011
[50]

PhysioBank, https://physionet.org/physiobank/

work page
[51]

Japan University Network Earthquake Catalog, http://wwweic.eri.u-tokyo.ac.jp/CATALOG/junec/

work page
[52]

D. R. Hunter and K. Lange, A Tutorial on MM Algo- rithms, The American Statistician 58, 30 (2004)

work page 2004

[1] [1]

At the initial time step s = 0, each event makes its own burst of size one

work page

[2] [2]

At each time step s, two bursts of sizes b and b′ are chosen at random with a probability proportional to Kbb′ [Eq. (5)]

work page

[3] [3]

Two merged bursts are assigned to left and right children nodes at random

Those two bursts are merged to make another burst of size b + b′, which becomes a parent node. Two merged bursts are assigned to left and right children nodes at random

work page

[4] [4]

Once the ordinal burst tree is generated, we randomly draw n − 1 IETs from the power-law IET distribution as P (τ) = τ −α Pτc τ =1 τ −α for τ = 1,

Steps 2–3 are repeated until all events belong to the single burst of size n. Once the ordinal burst tree is generated, we randomly draw n − 1 IETs from the power-law IET distribution as P (τ) = τ −α Pτc τ =1 τ −α for τ = 1, . . . , τc, (15) where α is the power-law exponent and τc is the upper bound. These IETs are sorted in a descending order to obtain ...

work page

[5] [5]

Barab´ asi and M

A.-L. Barab´ asi and M. P´ osfai,Network Science (Cam- bridge University Press, Cambridge, 2016)

work page 2016

[6] [6]

M. E. J. Newman, Networks, second edition ed. (Oxford University Press, Oxford, United Kingdom; New York, NY, United States of America, 2018)

work page 2018

[7] [7]

Menczer, S

F. Menczer, S. Fortunato, and C. A. Davis, A First Course in Network Science (Cambridge University Press, Cambridge, 2020)

work page 2020

[8] [8]

S. N. Dorogovtsev and J. F. F. Mendes, The Nature of Complex Networks (Oxford University Press, Oxford, UK; New York, NY, 2022)

work page 2022

[9] [9]

Easley and J

D. Easley and J. Kleinberg, Networks, Crowds, and Mar- kets: Reasoning About a Highly Connected World (Cam- bridge University Press, 2010)

work page 2010

[10] [10]

Barrat, M

A. Barrat, M. Barth´ elemy, and A. Vespignani,Dynami- cal Processes on Complex Networks , 1st ed. (Cambridge University Press, 2008)

work page 2008

[11] [11]

Holme and J

P. Holme and J. Saram¨ aki, Temporal networks, Physics Reports 519, 97 (2012)

work page 2012

[12] [12]

Holme and J

P. Holme and J. Saram¨ aki, eds.,Temporal Network The- ory, Computational Social Sciences (Springer Interna- tional Publishing, Cham, 2019)

work page 2019

[13] [13]

Masuda and R

N. Masuda and R. Lambiotte, A Guide to Temporal Net- works, Series on Complexity Science (World Scientific, New Jersey, 2016)

work page 2016

[14] [14]

Barab´ asi, The origin of bursts and heavy tails in human dynamics, Nature 435, 207 (2005)

A.-L. Barab´ asi, The origin of bursts and heavy tails in human dynamics, Nature 435, 207 (2005)

work page 2005

[15] [15]

Karsai, H.-H

M. Karsai, H.-H. Jo, and K. Kaski, Bursty Human Dy- namics (Springer International Publishing, Cham, 2018)

work page 2018

[16] [16]

Jo and T

H.-H. Jo and T. Hiraoka, Bursty Time Series Analysis for Temporal Networks, in Temporal Network Theory , edited by P. Holme and J. Saram¨ aki (Springer Interna- tional Publishing, Cham, 2023) 2nd ed., pp. 165–183

work page 2023

[17] [17]

P. Bak, C. Tang, and K. Wiesenfeld, Self-organized crit- icality: An explanation of the 1/f noise, Physical Review Letters 59, 381 (1987)

work page 1987

[18] [18]

H. J. Jensen, Self-Organized Criticality: Emergent Com- plex Behavior in Physical and Biological Systems , 1st ed. (Cambridge University Press, 1998)

work page 1998

[19] [19]

M. S. Wheatland, P. A. Sturrock, and J. M. McTiernan, The Waiting-Time Distribution of Solar Flare Hard X- Ray Bursts, The Astrophysical Journal 509, 448 (1998)

work page 1998

[20] [20]

Corral, Long-term clustering, scaling, and universal- ity in the temporal occurrence of earthquakes, Physical Review Letters 92, 108501 (2004)

´A. Corral, Long-term clustering, scaling, and universal- ity in the temporal occurrence of earthquakes, Physical Review Letters 92, 108501 (2004)

work page 2004

[21] [21]

J. M. Beggs and D. Plenz, Neuronal Avalanches in Neo- cortical Circuits, The Journal of Neuroscience 23, 11167 (2003)

work page 2003

[22] [22]

Petermann, T

T. Petermann, T. C. Thiagarajan, M. A. Lebedev, M. A. L. Nicolelis, D. R. Chialvo, and D. Plenz, Spon- taneous cortical activity in awake monkeys composed of neuronal avalanches, Proceedings of the National Academy of Sciences 106, 15921 (2009)

work page 2009

[23] [23]

Kemuriyama, H

T. Kemuriyama, H. Ohta, Y. Sato, S. Maruyama, M. Tandai-Hiruma, K. Kato, and Y. Nishida, A power- law distribution of inter-spike intervals in renal sym- pathetic nerve activity in salt-sensitive hypertension- induced chronic heart failure, BioSystems 101, 144 (2010)

work page 2010

[24] [24]

Goh and A.-L

K.-I. Goh and A.-L. Barab´ asi, Burstiness and memory in complex systems, EPL (Europhysics Letters) 81, 48002 (2008)

work page 2008

[25] [25]

Crane and D

R. Crane and D. Sornette, Robust dynamic classes re- vealed by measuring the response function of a social system, Proceedings of the National Academy of Sciences 105, 15649 (2008)

work page 2008

[26] [26]

Rybski, S

D. Rybski, S. V. Buldyrev, S. Havlin, F. Liljeros, and H. A. Makse, Scaling laws of human interaction activity, Proceedings of the National Academy of Sciences 106, 12640 (2009)

work page 2009

[27] [27]

H.-H. Jo, T. Hiraoka, and M. Kivel¨ a, Burst-tree decom- position of time series reveals the structure of temporal correlations, Scientific Reports 10, 12202 (2020)

work page 2020

[28] [28]

J. Choi, T. Hiraoka, and H.-H. Jo, Individual-driven versus interaction-driven burstiness in human dynamics: The case of Wikipedia edit history, Physical Review E 104, 014312 (2021). 8

work page 2021

[29] [29]

Karsai, K

M. Karsai, K. Kaski, A.-L. Barab´ asi, and J. Kert´ esz, Uni- versal features of correlated bursty behaviour, Scientific Reports 2, 397 (2012)

work page 2012

[30] [30]

Birhanu and H.-H

T. Birhanu and H.-H. Jo, Burst-tree structure and higher-order temporal correlations, Physical Review E 111, 014308 (2025)

work page 2025

[31] [31]

T. Pham, P. Sheridan, and H. Shimodaira, PAFit: A Sta- tistical Method for Measuring Preferential Attachment in Temporal Complex Networks, PLoS ONE 10, e0137796 (2015)

work page 2015

[32] [32]

Barab´ asi and R

A.-L. Barab´ asi and R. Albert, Emergence of Scaling in Random Networks, Science 286, 509 (1999)

work page 1999

[33] [33]

P. L. Krapivsky, S. Redner, and F. Leyvraz, Connectivity of Growing Random Networks, Physical Review Letters 85, 4629 (2000)

work page 2000

[34] [34]

M. E. J. Newman, Clustering and preferential attach- ment in growing networks, Physical Review E64, 025102 (2001)

work page 2001

[35] [35]

Jeong, Z

H. Jeong, Z. N´ eda, and A. L. Barab´ asi, Measuring prefer- ential attachment in evolving networks, Europhysics Let- ters (EPL) 61, 567 (2003)

work page 2003

[36] [36]

Sheridan, Y

P. Sheridan, Y. Yagahara, and H. Shimodaira, Measur- ing preferential attachment in growing networks with missing-timelines using Markov chain Monte Carlo, Phys- ica A: Statistical Mechanics and its Applications 391, 5031 (2012)

work page 2012

[37] [37]

Kunegis, M

J. Kunegis, M. Blattner, and C. Moser, Preferential at- tachment in online networks: Measurement and explana- tions, in Proceedings of the 5th Annual ACM Web Science Conference (ACM, Paris France, 2013) pp. 205–214

work page 2013

[38] [38]

W. H. Stockmayer, Theory of Molecular Size Distribu- tion and Gel Formation in Branched-Chain Polymers, The Journal of Chemical Physics 11, 45 (1943)

work page 1943

[39] [39]

Lushnikov, Evolution of coagulating systems, Journal of Colloid and Interface Science 45, 549 (1973)

A. Lushnikov, Evolution of coagulating systems, Journal of Colloid and Interface Science 45, 549 (1973)

work page 1973

[40] [40]

W. H. White, On the form of steady-state solutions to the coagulation equations, Journal of Colloid and Interface Science 87, 204 (1982)

work page 1982

[41] [41]

E. M. Hendriks, M. H. Ernst, and R. M. Ziff, Coagulation equations with gelation, Journal of Statistical Physics31, 519 (1983)

work page 1983

[42] [42]

D. J. Aldous, Deterministic and Stochastic Models for Coalescence (Aggregation and Coagulation): A Review of the Mean-Field Theory for Probabilists, Bernoulli 5, 3 (1999)

work page 1999

[43] [43]

M. H. Lee, A survey of numerical solutions to the coagu- lation equation, Journal of Physics A: Mathematical and General 34, 10219 (2001)

work page 2001

[44] [44]

Leyvraz, Scaling theory and exactly solved models in the kinetics of irreversible aggregation, Physics Reports 383, 95 (2003)

F. Leyvraz, Scaling theory and exactly solved models in the kinetics of irreversible aggregation, Physics Reports 383, 95 (2003)

work page 2003

[45] [45]

Leyvraz, Rigorous Results in the Scaling Theory of Irreversible Aggregation Kinetics:, Journal of Nonlinear Mathematical Physics 12, 449 (2005)

F. Leyvraz, Rigorous Results in the Scaling Theory of Irreversible Aggregation Kinetics:, Journal of Nonlinear Mathematical Physics 12, 449 (2005)

work page 2005

[46] [46]

J. A. Wattis, An introduction to mathematical models of coagulation–fragmentation processes: A discrete deter- ministic mean-field approach, Physica D: Nonlinear Phe- nomena 222, 1 (2006)

work page 2006

[47] [47]

Birhanu and H.-H

T. Birhanu and H.-H. Jo, Codes for the maximum likelihood estimation method of burst-merging kernels, https://github.com/tibebe22/Kernel-estimate (2025)

work page 2025

[48] [48]

English Wikipedia, https://dumps.wikimedia.org/

work page

[49] [49]

Yang and J

J. Yang and J. Leskovec, Patterns of temporal variation in online media, in Proceedings of the Fourth ACM In- ternational Conference on Web Search and Data Mining (ACM Press, Hong Kong, China, 2011) pp. 177–186

work page 2011

[50] [50]

PhysioBank, https://physionet.org/physiobank/

work page

[51] [51]

Japan University Network Earthquake Catalog, http://wwweic.eri.u-tokyo.ac.jp/CATALOG/junec/

work page

[52] [52]

D. R. Hunter and K. Lange, A Tutorial on MM Algo- rithms, The American Statistician 58, 30 (2004)

work page 2004