Multitaper Spectral Analysis of Neuronal Spiking Activity Driven by Latent Stationary Processes

Behtash Babadi; Proloy Das

arxiv: 1906.08451 · v1 · pith:VRBJOBJEnew · submitted 2019-06-20 · 📡 eess.SP · cs.IT· math.IT· q-bio.NC· stat.ME

Multitaper Spectral Analysis of Neuronal Spiking Activity Driven by Latent Stationary Processes

Proloy Das , Behtash Babadi This is my paper

Pith reviewed 2026-05-25 19:49 UTC · model grok-4.3

classification 📡 eess.SP cs.ITmath.ITq-bio.NCstat.ME

keywords multitaper spectral analysispoint processspiking activitylatent processesmaximum likelihoodeigen-spectraneural covariatesbias-variance trade-off

0 comments

The pith

A multitaper spectral estimator for spiking data infers latent process eigen-spectra directly from auxiliary statistics using maximum likelihood.

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

The paper develops a variant of the multitaper method for estimating spectral properties of neural covariates from spiking activity. Standard approaches for continuous signals do not address the nonlinearities inherent in point process observations. The authors construct auxiliary spiking statistics that support maximum likelihood recovery of the eigen-spectra of the underlying latent stationary processes. This construction allows direct and efficient computation of the multitaper estimate. Simulated data comparisons indicate gains in the bias-variance trade-off relative to prior techniques.

Core claim

By constructing auxiliary spiking statistics from point process data, the eigen-spectra of the underlying latent stationary processes can be directly inferred using maximum likelihood estimation, allowing the multitaper estimate to be efficiently computed with significant gains in the bias-variance trade-off.

What carries the argument

Auxiliary spiking statistics that permit direct maximum likelihood inference of eigen-spectra for the multitaper method.

If this is right

The method supports spectral analysis of neural covariates from spiking datasets.
It achieves better bias-variance trade-off in spectral estimation.
It is applicable to data driven by latent stationary processes.
Comparison to existing methods shows improved performance on simulated data.

Where Pith is reading between the lines

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

Researchers could apply this to real neural recordings to study cognitive functions without additional nonlinearity modeling.
The approach might extend to other point process spectral problems in neuroscience.
Direct recovery suggests potential for parameter-free derivations in related analyses.

Load-bearing premise

The latent neural covariates are driven by stationary processes whose eigen-spectra can be recovered directly from auxiliary spiking statistics via maximum likelihood without additional modeling assumptions on the point-process nonlinearity.

What would settle it

A simulation where the maximum likelihood estimates from the auxiliary statistics do not accurately recover the known eigen-spectra of the latent process would falsify the recovery claim.

Figures

Figures reproduced from arXiv: 1906.08451 by Behtash Babadi, Proloy Das.

**Figure 2.** Figure 2: (A) A snapshot of the simulated AR process for 200 ≤ k ≤ 350. (B) Raster plot of the corresponding neuronal ensemble activity. existing methods: (1) PSTH-PSD, where the PSD is computed by forming the MTM estimate of the ensemble peristimulus time histogram (PSTH), i.e., the average spike trains, and (2) SS-PSD, where xk is first estimated using a state-space model xk = xk−1 + wk, followed by forming its MT… view at source ↗

**Figure 3.** Figure 3: Comparison of the PSD estimates. (A) PMTM, PSTH-PSD, SS-PSD, Oracle PSD (all using α = 5, J = 8), and the true PSD. (B) PMTM estimates for L = 5, 10, 15, and 20. Fig. 3A shows the PMTM (black), PSTH-PSD (green), SS-PSD (aqua) and the true PSD (blue) for the realization shown in 2 in log-scale. For comparison purposes, we have also included the MTM PSD estimate 8 [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗

read the original abstract

Investigating the spectral properties of the neural covariates that underlie spiking activity is an important problem in systems neuroscience, as it allows to study the role of brain rhythms in cognitive functions. While the spectral estimation of continuous time-series is a well-established domain, computing the spectral representation of these neural covariates from spiking data sets forth various challenges due to the intrinsic non-linearities involved. In this paper, we address this problem by proposing a variant of the multitaper method specifically tailored for point process data. To this end, we construct auxiliary spiking statistics from which the eigen-spectra of the underlying latent process can be directly inferred using maximum likelihood estimation, and thereby the multitaper estimate can be efficiently computed. Comparison of our proposed technique to existing methods using simulated data reveals significant gains in terms of the bias-variance trade-off.

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 a multitaper variant for spike trains via auxiliary stats and MLE, but the direct recovery step looks vulnerable to the unknown nonlinearity.

read the letter

The main takeaway is a multitaper spectral estimator built for point-process spiking data. They define auxiliary spiking statistics, then use maximum likelihood to recover the eigen-spectra of the latent stationary process before forming the multitaper estimate. That construction is the actual novelty; it is not just a routine extension of the usual continuous-time method. The abstract reports better bias-variance performance than existing approaches on simulated data, which is the concrete evidence offered. That is useful to note because the problem itself is practical in systems neuroscience. Standard multitaper does not apply cleanly to spikes, so a targeted fix has a clear audience. The soft spot is the load-bearing claim that the auxiliary statistics permit direct MLE recovery without additional modeling assumptions on the point-process nonlinearity. If the intensity function is nonlinear, the second-order statistics of the observed spikes are not a simple function of the latent spectrum, so the likelihood would normally depend on the link. The abstract asserts the recovery is direct but supplies neither the explicit form of the auxiliaries nor the derivation showing the score equations stay independent of that link. The simulations are cited without details on the generative model, trial counts, or exclusion criteria, so it is difficult to judge how much the reported gains rely on favorable conditions. This work is aimed at researchers who already do spectral analysis on spike trains and need a practical tool for latent rhythms. A reader in that niche could extract value from the technique once the assumptions are verified. I would send it to peer review; the idea is specific enough and the target application narrow enough that referees can check the math and the simulation setup without wasting broad effort.

Referee Report

2 major / 0 minor

Summary. The paper proposes a variant of the multitaper method for spectral estimation of latent stationary processes that drive neuronal spiking activity. It constructs auxiliary spiking statistics from which the eigen-spectra of the latent process are recovered via maximum likelihood estimation, after which the multitaper estimate is computed. Simulations are reported to show gains in the bias-variance trade-off relative to existing methods.

Significance. If the central recovery step holds without hidden dependence on the point-process nonlinearity, the method would offer a useful adaptation of classical multitaper techniques to point-process data, with potential value for analyzing brain rhythms in systems neuroscience.

major comments (2)

[Abstract] Abstract: the claim that eigen-spectra 'can be directly inferred using maximum likelihood estimation' from auxiliary spiking statistics 'without additional modeling assumptions' on the point-process nonlinearity is load-bearing for the central contribution. The likelihood formed from any auxiliary statistic (binned counts, inter-spike intervals, etc.) is obtained by integrating the latent Gaussian process through an unknown intensity function; unless the score equations are shown to be independent of the link function, the MLE implicitly requires a parametric model for the nonlinearity or an explicit proof of invariance, neither of which appears in the given text.
[Abstract] Abstract (simulation results): the reported 'significant gains in terms of the bias-variance trade-off' are presented without derivation details, error bars, exclusion criteria, or the precise definition of the auxiliary statistics and likelihood, rendering the empirical support difficult to evaluate.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We are grateful to the referee for their insightful comments on our manuscript. Below we provide point-by-point responses to the major comments.

read point-by-point responses

Referee: [Abstract] Abstract: the claim that eigen-spectra 'can be directly inferred using maximum likelihood estimation' from auxiliary spiking statistics 'without additional modeling assumptions' on the point-process nonlinearity is load-bearing for the central contribution. The likelihood formed from any auxiliary statistic (binned counts, inter-spike intervals, etc.) is obtained by integrating the latent Gaussian process through an unknown intensity function; unless the score equations are shown to be independent of the link function, the MLE implicitly requires a parametric model for the nonlinearity or an explicit proof of invariance, neither of which appears in the given text.

Authors: The referee correctly identifies that the abstract's claim regarding inference without additional modeling assumptions is central. While the full manuscript details the construction of auxiliary statistics and the MLE procedure for recovering the eigen-spectra of the latent process, it does not include an explicit demonstration that the score equations are independent of the link function. We will therefore revise the paper to add this proof in a new appendix or section, ensuring the claim is rigorously supported. revision: yes
Referee: [Abstract] Abstract (simulation results): the reported 'significant gains in terms of the bias-variance trade-off' are presented without derivation details, error bars, exclusion criteria, or the precise definition of the auxiliary statistics and likelihood, rendering the empirical support difficult to evaluate.

Authors: We acknowledge that the simulation results section would benefit from more comprehensive details to allow full evaluation. In the revised version, we will provide the precise definitions of the auxiliary statistics and likelihood, include derivation details for the bias-variance analysis, add error bars to the figures, and specify any exclusion criteria. This will enhance the transparency and reproducibility of the empirical findings. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper's derivation constructs auxiliary spiking statistics and applies maximum likelihood estimation to recover eigen-spectra of a latent stationary process before computing a multitaper estimate. This chain is presented as a novel methodological step for point-process data, with performance evaluated via external simulation benchmarks rather than internal fits or self-citations. No quoted equations or steps reduce a claimed prediction to a fitted parameter by construction, nor does any load-bearing premise collapse to a self-citation chain or ansatz smuggled from prior author work. The method therefore remains self-contained against the provided description.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the existence of latent stationary processes whose spectra are recoverable from auxiliary statistics; no free parameters or invented entities are mentioned in the abstract.

axioms (1)

domain assumption Neural covariates are driven by latent stationary processes.
Stated in the title and abstract as the modeling premise for the spectral analysis.

pith-pipeline@v0.9.0 · 5680 in / 1176 out tokens · 18012 ms · 2026-05-25T19:49:17.193344+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

33 extracted references · 33 canonical work pages

[1]

T. F. Quatieri, Discrete-time Speech Signal Processing : Principles and Practice, Prentice Hall, 2008

work page 2008
[2]

J. S. Lim, Two-dimensional signal and image processing, Englewood Cliﬀs, NJ, Prentice Hall, 1990, 710 p

work page 1990
[3]

Buzsaki, Rhythms of the Brain, Oxford University Pres s, 2009

G. Buzsaki, Rhythms of the Brain, Oxford University Pres s, 2009

work page 2009
[4]

D. J. Thomson, Spectrum estimation and harmonic analysi s, Proceedings of the IEEE 70 (9) (1982) 1055–1096. doi:10.1109/PROC.1982.12433

work page doi:10.1109/proc.1982.12433 1982
[5]

W alden, A uniﬁed view of multitaper multivariate spec tral estimation, Biometrika 87 (4) (2000) 767–788

A. W alden, A uniﬁed view of multitaper multivariate spec tral estimation, Biometrika 87 (4) (2000) 767–788

work page 2000
[6]

Babadi, E

B. Babadi, E. N. Brown, A review of multitaper spectral an alysis, IEEE Transactions on Biomedical Engineering 61 (5) (2014) 1555–1564. doi:10.1109/TBME.2014.2311996

work page doi:10.1109/tbme.2014.2311996 2014
[7]

Alarcon, C

G. Alarcon, C. D. Binnie, R. D. Elwes, C. E. Polkey, Power s pectrum and intracranial EEG patterns at seizure onset in partial epilepsy, Electroencephalography and clinical ne urophysiology 94 5 (1995) 326–37

work page 1995
[8]

Fisher, W

R. Fisher, W. W ebber, R. Lesser, S. Arroyo, S. Uematsu, Hi gh-frequency EEG activity at the start of seizures, Journal of Clinical Neurophysiology 9 (3) (1992) 441–448

work page 1992
[9]

L. M. W ard, Synchronous neural oscillations and cogniti ve processes, Trends in cognitive sciences 7 (12) (2003) 553 –559

work page 2003
[10]

P. Das, B. Babadi, Dynamic bayesian multitaper spectra l analysis, IEEE Transactions on Signal Processing 66 (6) (2 018) 1394–1409. doi:10.1109/TSP.2017.2787146

work page doi:10.1109/tsp.2017.2787146 2017
[11]

S.-E. Kim, M. K. Behr, D. Ba, E. N. Brown, State-space mul titaper time-frequency analysis, Proceedings of the Natio nal Academy of Sciences doi:10.1073/pnas.1702877115. URL http://www.pnas.org/content/early/2017/12/15/1702877115

work page doi:10.1073/pnas.1702877115 2017
[12]

Chalk, J

M. Chalk, J. L. Herrero, M. A. Gieselmann, L. S. Delicato , S. Gotthardt, A. Thiele, Attention reduces stimulus-driv en gamma frequency oscillations and spike ﬁeld coherence in V1 , Neuron 66 (1) (2010) 114 – 125

work page 2010
[13]

B. C. Lewandowski, M. Schmidt, Short bouts of vocalizat ion induce long-lasting fast gamma oscillations in a sensor imotor nucleus, The Journal of Neuroscience 31 (39) (2011) 13936–4 8

work page 2011
[14]

I. M. Park, S. Seth, A. R. Paiva, L. Li, J. C. Principe, Ker nel methods on spike train space for neuroscience: a tutoria l, IEEE Signal Processing Magazine 30 (4) (2013) 149–160

work page 2013
[15]

Truccolo, U

W. Truccolo, U. T. Eden, M. R. Fellows, J. P. Donoghue, E. N. Brown, A point process framework for relating neural spiking activity to spiking history, neural ensemble, and e xtrinsic covariate eﬀects, Journal of Neurophysiology 93 ( 2) (2005) 1074–1089. doi:10.1152/jn.00697.2004

work page doi:10.1152/jn.00697.2004 2005
[16]

Paninski, Maximum likelihood estimation of cascade point-process neural encoding models, Computat

L. Paninski, Maximum likelihood estimation of cascade point-process neural encoding models, Computat. Neural Sy st. 15 (2004) 243–262

work page 2004
[17]

A. W u, N. G. Roy, S. Keeley, J. W. Pillow, Gaussian proces s based nonlinear latent structure discovery in multivaria te spike train data, in: Advances in Neural Information Proces sing Systems, 2017, pp. 3496–3505

work page 2017
[18]

S. Xu, Y. Li, T. Huang, R. H. Chan, A sparse multiwavelet- based generalized laguerre–volterra model for identifyin g time-varying neural dynamics from spiking activities, Ent ropy 19 (8) (2017) 425

work page 2017
[19]

A. C. Smith, E. N. Brown, Estimating a state-space model from point process observations, Neural Computation 15 (20 03) 965–991. 10

work page
[20]

P. M. Djuric, M. Vemula, M. F. Bugallo, Target tracking b y particle ﬁltering in binary sensor networks, IEEE Transac tions on Signal Processing 56 (6) (2008) 2229–2238

work page 2008
[21]

E. N. Brown, R. E. Kass, P. P. Mitra, Multiple neural spik e train data analysis: state-of-the-art and future challen ges, Nature neuroscience 7 (2004) 456–461

work page 2004
[22]

D. B. Percival, A. T. W alden, Spectral Analysis for Phys ical Applications, Cambridge University Press, 1993

work page 1993
[23]

Miran, P

S. Miran, P. L. Purdon, E. N. Brown, B. Babadi, Robust est imation of sparse narrowband spectra from neuronal spiking data, IEEE Transactions on Biomedical Engineering 64 (10) ( 2017) 2462–2474

work page 2017
[24]

L. D. Lewis, V. S. W einer, E. A. Mukamel, J. A. Donoghue, E . N. Eskandar, J. R. Madsen, W. S. Anderson, L. R. Hochberg, S. S. Cash, E. N. Brown, P. L. Purdon, Rapid fragmen tation of neuronal networks at the onset of propofol- induced unconsciousness, Proceedings of the National Acad emy of Sciences 109 (49) (2012) E3377–E3386

work page 2012
[25]

Sheikhattar, J

A. Sheikhattar, J. B. Fritz, S. A. Shamma, B. Babadi, Rec ursive sparse point process regression with application to spectrotemporal receptive ﬁeld plasticity analysis, IEEE Transactions on Signal Processing 64 (8) (2016) 2026–2039

work page 2016
[26]

Loeve, Probability Theory, D

M. Loeve, Probability Theory, D. Van Nostrand Co., Lond on, 1963

work page 1963
[27]

Slepian, Prolate spheroidal wave functions, Fourie r analysis, and uncertainty-V: the discrete case, Bell Syst

D. Slepian, Prolate spheroidal wave functions, Fourie r analysis, and uncertainty-V: the discrete case, Bell Syst . Tech. J. 57 (5) (1978) 1371–1430. doi:10.1002/j.1538-7305.1978.tb02104.x

work page doi:10.1002/j.1538-7305.1978.tb02104.x 1978
[28]

D. R. Cox, V. Isham, Point Processes, Chapman and Hall, 1 980

work page
[29]

Billingsley, Probability and measure, John Wiley & S ons, 2008

P. Billingsley, Probability and measure, John Wiley & S ons, 2008

work page 2008
[30]

A. P. Dempster, N. M. Laird, D. B. Rubin, Maximum likelih ood from incomplete data via the em algorithm, Journal of the royal statistical society. Series B (methodological) ( 1977) 1–38

work page 1977
[31]

S. Boyd, L. Vandenberghe, Convex optimization, Cambri dge university press, 2004

work page 2004
[32]

The Point Process Multitaper Method, Available on GitH ub Repository: https://github.com/proloyd/PMTM, 2018

work page 2018
[33]

Forli, D

A. Forli, D. Vecchia, N. Binini, F. Succol, S. Bovetti, C . Moretti, F. Nespoli, M. Mahn, C. A. Baker, M. M. Bolton, O. Yizhar, T. Fellin, Two-photon bidirectional control and imaging of neuronal excitability with high spatial resolut ion in vivo, Cell reports 22 (11) (2018) 3087–3098. 11

work page 2018

[1] [1]

T. F. Quatieri, Discrete-time Speech Signal Processing : Principles and Practice, Prentice Hall, 2008

work page 2008

[2] [2]

J. S. Lim, Two-dimensional signal and image processing, Englewood Cliﬀs, NJ, Prentice Hall, 1990, 710 p

work page 1990

[3] [3]

Buzsaki, Rhythms of the Brain, Oxford University Pres s, 2009

G. Buzsaki, Rhythms of the Brain, Oxford University Pres s, 2009

work page 2009

[4] [4]

D. J. Thomson, Spectrum estimation and harmonic analysi s, Proceedings of the IEEE 70 (9) (1982) 1055–1096. doi:10.1109/PROC.1982.12433

work page doi:10.1109/proc.1982.12433 1982

[5] [5]

W alden, A uniﬁed view of multitaper multivariate spec tral estimation, Biometrika 87 (4) (2000) 767–788

A. W alden, A uniﬁed view of multitaper multivariate spec tral estimation, Biometrika 87 (4) (2000) 767–788

work page 2000

[6] [6]

Babadi, E

B. Babadi, E. N. Brown, A review of multitaper spectral an alysis, IEEE Transactions on Biomedical Engineering 61 (5) (2014) 1555–1564. doi:10.1109/TBME.2014.2311996

work page doi:10.1109/tbme.2014.2311996 2014

[7] [7]

Alarcon, C

G. Alarcon, C. D. Binnie, R. D. Elwes, C. E. Polkey, Power s pectrum and intracranial EEG patterns at seizure onset in partial epilepsy, Electroencephalography and clinical ne urophysiology 94 5 (1995) 326–37

work page 1995

[8] [8]

Fisher, W

R. Fisher, W. W ebber, R. Lesser, S. Arroyo, S. Uematsu, Hi gh-frequency EEG activity at the start of seizures, Journal of Clinical Neurophysiology 9 (3) (1992) 441–448

work page 1992

[9] [9]

L. M. W ard, Synchronous neural oscillations and cogniti ve processes, Trends in cognitive sciences 7 (12) (2003) 553 –559

work page 2003

[10] [10]

P. Das, B. Babadi, Dynamic bayesian multitaper spectra l analysis, IEEE Transactions on Signal Processing 66 (6) (2 018) 1394–1409. doi:10.1109/TSP.2017.2787146

work page doi:10.1109/tsp.2017.2787146 2017

[11] [11]

S.-E. Kim, M. K. Behr, D. Ba, E. N. Brown, State-space mul titaper time-frequency analysis, Proceedings of the Natio nal Academy of Sciences doi:10.1073/pnas.1702877115. URL http://www.pnas.org/content/early/2017/12/15/1702877115

work page doi:10.1073/pnas.1702877115 2017

[12] [12]

Chalk, J

M. Chalk, J. L. Herrero, M. A. Gieselmann, L. S. Delicato , S. Gotthardt, A. Thiele, Attention reduces stimulus-driv en gamma frequency oscillations and spike ﬁeld coherence in V1 , Neuron 66 (1) (2010) 114 – 125

work page 2010

[13] [13]

B. C. Lewandowski, M. Schmidt, Short bouts of vocalizat ion induce long-lasting fast gamma oscillations in a sensor imotor nucleus, The Journal of Neuroscience 31 (39) (2011) 13936–4 8

work page 2011

[14] [14]

I. M. Park, S. Seth, A. R. Paiva, L. Li, J. C. Principe, Ker nel methods on spike train space for neuroscience: a tutoria l, IEEE Signal Processing Magazine 30 (4) (2013) 149–160

work page 2013

[15] [15]

Truccolo, U

W. Truccolo, U. T. Eden, M. R. Fellows, J. P. Donoghue, E. N. Brown, A point process framework for relating neural spiking activity to spiking history, neural ensemble, and e xtrinsic covariate eﬀects, Journal of Neurophysiology 93 ( 2) (2005) 1074–1089. doi:10.1152/jn.00697.2004

work page doi:10.1152/jn.00697.2004 2005

[16] [16]

Paninski, Maximum likelihood estimation of cascade point-process neural encoding models, Computat

L. Paninski, Maximum likelihood estimation of cascade point-process neural encoding models, Computat. Neural Sy st. 15 (2004) 243–262

work page 2004

[17] [17]

A. W u, N. G. Roy, S. Keeley, J. W. Pillow, Gaussian proces s based nonlinear latent structure discovery in multivaria te spike train data, in: Advances in Neural Information Proces sing Systems, 2017, pp. 3496–3505

work page 2017

[18] [18]

S. Xu, Y. Li, T. Huang, R. H. Chan, A sparse multiwavelet- based generalized laguerre–volterra model for identifyin g time-varying neural dynamics from spiking activities, Ent ropy 19 (8) (2017) 425

work page 2017

[19] [19]

A. C. Smith, E. N. Brown, Estimating a state-space model from point process observations, Neural Computation 15 (20 03) 965–991. 10

work page

[20] [20]

P. M. Djuric, M. Vemula, M. F. Bugallo, Target tracking b y particle ﬁltering in binary sensor networks, IEEE Transac tions on Signal Processing 56 (6) (2008) 2229–2238

work page 2008

[21] [21]

E. N. Brown, R. E. Kass, P. P. Mitra, Multiple neural spik e train data analysis: state-of-the-art and future challen ges, Nature neuroscience 7 (2004) 456–461

work page 2004

[22] [22]

D. B. Percival, A. T. W alden, Spectral Analysis for Phys ical Applications, Cambridge University Press, 1993

work page 1993

[23] [23]

Miran, P

S. Miran, P. L. Purdon, E. N. Brown, B. Babadi, Robust est imation of sparse narrowband spectra from neuronal spiking data, IEEE Transactions on Biomedical Engineering 64 (10) ( 2017) 2462–2474

work page 2017

[24] [24]

L. D. Lewis, V. S. W einer, E. A. Mukamel, J. A. Donoghue, E . N. Eskandar, J. R. Madsen, W. S. Anderson, L. R. Hochberg, S. S. Cash, E. N. Brown, P. L. Purdon, Rapid fragmen tation of neuronal networks at the onset of propofol- induced unconsciousness, Proceedings of the National Acad emy of Sciences 109 (49) (2012) E3377–E3386

work page 2012

[25] [25]

Sheikhattar, J

A. Sheikhattar, J. B. Fritz, S. A. Shamma, B. Babadi, Rec ursive sparse point process regression with application to spectrotemporal receptive ﬁeld plasticity analysis, IEEE Transactions on Signal Processing 64 (8) (2016) 2026–2039

work page 2016

[26] [26]

Loeve, Probability Theory, D

M. Loeve, Probability Theory, D. Van Nostrand Co., Lond on, 1963

work page 1963

[27] [27]

Slepian, Prolate spheroidal wave functions, Fourie r analysis, and uncertainty-V: the discrete case, Bell Syst

D. Slepian, Prolate spheroidal wave functions, Fourie r analysis, and uncertainty-V: the discrete case, Bell Syst . Tech. J. 57 (5) (1978) 1371–1430. doi:10.1002/j.1538-7305.1978.tb02104.x

work page doi:10.1002/j.1538-7305.1978.tb02104.x 1978

[28] [28]

D. R. Cox, V. Isham, Point Processes, Chapman and Hall, 1 980

work page

[29] [29]

Billingsley, Probability and measure, John Wiley & S ons, 2008

P. Billingsley, Probability and measure, John Wiley & S ons, 2008

work page 2008

[30] [30]

A. P. Dempster, N. M. Laird, D. B. Rubin, Maximum likelih ood from incomplete data via the em algorithm, Journal of the royal statistical society. Series B (methodological) ( 1977) 1–38

work page 1977

[31] [31]

S. Boyd, L. Vandenberghe, Convex optimization, Cambri dge university press, 2004

work page 2004

[32] [32]

The Point Process Multitaper Method, Available on GitH ub Repository: https://github.com/proloyd/PMTM, 2018

work page 2018

[33] [33]

Forli, D

A. Forli, D. Vecchia, N. Binini, F. Succol, S. Bovetti, C . Moretti, F. Nespoli, M. Mahn, C. A. Baker, M. M. Bolton, O. Yizhar, T. Fellin, Two-photon bidirectional control and imaging of neuronal excitability with high spatial resolut ion in vivo, Cell reports 22 (11) (2018) 3087–3098. 11

work page 2018