pith. sign in

arxiv: 2607.01082 · v1 · pith:7NP3HW6Cnew · submitted 2026-07-01 · 💻 cs.LG

When Context Compensates for Sparse Event History: AlphaEarth for Spatio-Temporal Point-Process Forecasting

Yahya Aalaila , Mouad Elhamdi , Gerrit Gro{\ss}mann , Daniel Jenson , Elizaveta Semenova , Sebastian Vollmer This is my paper

Pith reviewed 2026-07-02 15:48 UTC · model grok-4.3

classification 💻 cs.LG
keywords spatio-temporal point processesAlphaEarth embeddingslog-Gaussian Cox processemergency medical services forecastingspatial contextsparse event historyout-of-region generalizationforecast transfer
0
0 comments X

The pith

AlphaEarth embeddings improve out-of-region point-process forecasts by 2-6 times when local event histories are only 1-2 weeks long.

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

The paper tests whether exogenous spatial context can substitute for scarce local event data when forecasting spatio-temporal point processes in new regions. It fixes a log-Gaussian Cox process backbone and compares an event-only version against the same model that receives AlphaEarth embeddings as additional linear spatial covariates. Experiments use emergency medical services data, eight held-out regions, fixed forecast anchors, and a sweep of history lengths from one week to two years, with all embeddings taken strictly before each anchor. Performance lifts appear in every regime, but the multiplicative improvement is largest at the shortest histories and shrinks as more local data becomes available. The setup isolates the contribution of the spatial context to generalization under data scarcity.

Core claim

When local event histories are sparse, adding pre-anchor AlphaEarth embeddings as linear spatial covariates to a log-Gaussian Cox process raises out-of-region predictive performance, with 2-6 times better results at 1-2 week histories that taper to 10-20 percent gains at 20-104 weeks.

What carries the argument

AlphaEarth embeddings supplied as exogenous linear spatial context in a fixed log-Gaussian Cox process backbone.

If this is right

  • Out-of-region performance improves across the full range of history lengths examined.
  • The largest multiplicative gains (2-6x) occur at the shortest histories of 1-2 weeks.
  • Gains shrink to 10-20 percent once histories reach 20 weeks or longer.
  • The fixed backbone isolates the effect of the added spatial context from other modeling decisions.
  • Contextual spatial information can stabilize forecasts transferred to new regions when local data is limited.

Where Pith is reading between the lines

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

  • The same compensation pattern may appear in other point-process domains such as crime incidents or seismic events when local records are short.
  • Testing the embeddings with a different backbone model while holding everything else fixed would separate the contribution of the context from the choice of process model.
  • Embeddings derived from other pre-trained spatial representations could be substituted to measure whether the compensation effect is specific to AlphaEarth.
  • The method could be applied to partially observed regions rather than fully held-out ones to check whether partial local history plus context yields further gains.

Load-bearing premise

That any measured performance difference is caused by the relevance and pre-anchor timing of the AlphaEarth embeddings rather than by other unstated choices in the two model variants.

What would settle it

Observing no difference in predictive scores between the event-only and context-augmented models across the same held-out regions and history lengths, or finding that the embeddings contain information from after the chosen forecast anchors.

Figures

Figures reproduced from arXiv: 2607.01082 by Daniel Jenson, Elizaveta Semenova, Gerrit Gro{\ss}mann, Mouad Elhamdi, Sebastian Vollmer, Yahya Aalaila.

Figure 1
Figure 1. Figure 1: Conceptual motivation. In spatial transfer, contextual [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Spatial distribution of EMS 911 calls in Montgomery [PITH_FULL_IMAGE:figures/full_fig_p003_2.png] view at source ↗
Figure 4
Figure 4. Figure 4: Held-out ELPD curves for two representative spatial masks. [PITH_FULL_IMAGE:figures/full_fig_p004_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Mean multiplicative improvement in held-out predictive [PITH_FULL_IMAGE:figures/full_fig_p004_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: Distribution of paired per-event log-score improvements [PITH_FULL_IMAGE:figures/full_fig_p005_6.png] view at source ↗
read the original abstract

Spatio-temporal point-process models must often generalise across space when local event histories are sparse. We study whether exogenous spatial context can compensate in such regimes. Using a fixed log-Gaussian Cox process backbone, we compare an event-only model with the same model augmented by AlphaEarth embeddings as linear spatial context. We evaluate spatial transfer on emergency medical services (EMS) forecasting across eight held-out regions, fixed forecast anchors, and a sweep over history length $w$, using only AlphaEarth (AE) embeddings available strictly before each anchor. AE improves out-of-region predictive performance across all history regimes, with the largest gains under scarce histories: approximately $2$--$6\times$ multiplicative improvements at $1-2$ weeks, tapering to roughly $10$--$20\%$ at $w=20$--$104$ weeks. These results show that contextual information can substantially stabilise spatially transferred point-process forecasts when event history is limited.

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.

Referee Report

2 major / 1 minor

Summary. The manuscript claims that augmenting a fixed log-Gaussian Cox process (LGCP) backbone with AlphaEarth embeddings as linear spatial context improves out-of-region predictive performance for spatio-temporal point-process forecasting on emergency medical services (EMS) data. Using strictly pre-anchor embeddings, it reports multiplicative gains of approximately 2--6 imes at history lengths w=1--2 weeks that taper to 10--20% at w=20--104 weeks across eight held-out regions.

Significance. If the central attribution holds under identical backbones and standard controls, the result would show that exogenous spatial embeddings can compensate for sparse local event histories in point-process models, providing a practical route to stabilize spatial transfer in data-limited forecasting settings.

major comments (2)
  1. [Abstract] Abstract: the abstract asserts a 'fixed log-Gaussian Cox process backbone' and 'the same model augmented' by AlphaEarth embeddings, yet supplies no explicit statement or verification that GP kernel hyperparameters, quadrature/MCMC settings, covariate scaling, optimization schedule, or intensity baseline are literally identical between the event-only and AE variants. This detail is load-bearing for attributing the reported 2--6 imes gains specifically to the embeddings rather than unstated implementation differences.
  2. [Abstract] Abstract: the reported multiplicative improvements lack any mention of the underlying predictive metric, cross-validation procedure, error bars, or controls for multiple testing and region selection, rendering it impossible to assess whether the gains survive standard statistical scrutiny.
minor comments (1)
  1. The abstract would be clearer if it named the exact evaluation metric (e.g., held-out log-likelihood or intensity RMSE) underlying the multiplicative factors.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the careful review and for identifying points that strengthen the clarity of our claims. We respond to each major comment below. Where the abstract can be improved without exceeding length limits, we will revise accordingly.

read point-by-point responses

  1. Referee: [Abstract] Abstract: the abstract asserts a 'fixed log-Gaussian Cox process backbone' and 'the same model augmented' by AlphaEarth embeddings, yet supplies no explicit statement or verification that GP kernel hyperparameters, quadrature/MCMC settings, covariate scaling, optimization schedule, or intensity baseline are literally identical between the event-only and AE variants. This detail is load-bearing for attributing the reported 2--6 times gains specifically to the embeddings rather than unstated implementation differences.

    Authors: The manuscript uses an identical LGCP backbone for both variants, with all listed components (kernel hyperparameters, quadrature/MCMC settings, covariate scaling, optimization schedule, and intensity baseline) held exactly fixed; this is verified in the Methods section. The abstract's phrasing 'fixed ... backbone' and 'the same model augmented' is intended to convey this, but we acknowledge the referee's point that an explicit verification statement would remove any ambiguity. We will revise the abstract to add: 'with all model hyperparameters and inference settings held fixed'. revision: yes

  2. Referee: [Abstract] Abstract: the reported multiplicative improvements lack any mention of the underlying predictive metric, cross-validation procedure, error bars, or controls for multiple testing and region selection, rendering it impossible to assess whether the gains survive standard statistical scrutiny.

    Authors: The metric is out-of-region log predictive density, evaluated on eight held-out regions with fixed forecast anchors and a sweep over history length w. Full results include error bars and region-level breakdowns. The abstract is space-constrained and cannot accommodate error bars or multiple-testing details, which belong in the main text. We will revise the abstract to specify the metric ('log predictive density') and the out-of-region transfer protocol. This addresses the core request while preserving the abstract's brevity. revision: partial

Circularity Check

0 steps flagged

No circularity; empirical comparison with fixed backbone is self-contained

full rationale

The paper reports an empirical evaluation of out-of-region predictive performance for a fixed LGCP backbone versus the same backbone augmented by strictly pre-anchor AlphaEarth embeddings as linear spatial covariates. The claimed gains (2-6x at short histories) are measured on held-out regions across a sweep of history lengths w and do not reduce to any fitted parameter being renamed as a prediction, any self-definitional relation, or a load-bearing self-citation chain. The derivation chain consists of standard point-process likelihood evaluation and spatial transfer metrics; no equation or result is equivalent to its inputs by construction.

Axiom & Free-Parameter Ledger

1 free parameters · 2 axioms · 0 invented entities

The central claim rests on the assumption that the chosen LGCP backbone plus linear embedding augmentation is a fair and stable comparator; no free parameters are explicitly named in the abstract, but the linear coefficients on the embeddings are implicitly fitted.

free parameters (1)
  • linear coefficients on AlphaEarth embeddings
    These are fitted within the augmented model to produce the reported performance lift.
axioms (2)
  • domain assumption The log-Gaussian Cox process is an appropriate fixed backbone for comparing event-only versus context-augmented models.
    The abstract states the backbone is held fixed but does not justify why this particular point-process family is suitable for EMS data.
  • domain assumption AlphaEarth embeddings are exogenous and available strictly before each forecast anchor.
    The abstract emphasizes pre-anchor availability but provides no verification details.

pith-pipeline@v0.9.1-grok · 5723 in / 1410 out tokens · 20674 ms · 2026-07-02T15:48:25.671022+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

75 extracted references · 11 canonical work pages · 2 internal anchors

  1. [1]

    arXiv preprint arXiv:2106.13097 , year=

    Understanding the spread of COVID-19 epidemic: A spatio-temporal point process view , author=. arXiv preprint arXiv:2106.13097 , year=

    work page arXiv

  2. [2]

    Journal of the Royal Statistical Society Series C: Applied Statistics , volume=

    Non-stationary spatio-temporal point process modeling for high-resolution COVID-19 data , author=. Journal of the Royal Statistical Society Series C: Applied Statistics , volume=. 2023 , publisher=

    2023

  3. [3]

    arXiv preprint arXiv:2409.10882 , year=

    Spatio-temporal-network point processes for modeling crime events with landmarks , author=. arXiv preprint arXiv:2409.10882 , year=

    work page arXiv

  4. [4]

    Spatial Statistics , volume=

    Prediction of intensity and location of seismic events using deep learning , author=. Spatial Statistics , volume=. 2021 , publisher=

    2021

  5. [5]

    Proceedings of the 21th ACM SIGKDD international conference on knowledge discovery and data mining , pages=

    Predicting ambulance demand: A spatio-temporal kernel approach , author=. Proceedings of the 21th ACM SIGKDD international conference on knowledge discovery and data mining , pages=

  6. [6]

    2003 , publisher=

    An introduction to the theory of point processes: volume I: elementary theory and methods , author=. 2003 , publisher=

    2003

  7. [7]

    2013 , publisher=

    Statistical analysis of spatial and spatio-temporal point patterns , author=. 2013 , publisher=

    2013

  8. [8]

    Biometrika , volume=

    Spectra of some self-exciting and mutually exciting point processes , author=. Biometrika , volume=. 1971 , publisher=

    1971

  9. [9]

    Statistical Science , volume=

    A review of self-exciting spatio-temporal point processes and their applications , author=. Statistical Science , volume=. 2018 , publisher=

    2018

  10. [10]

    Statistical Science , volume=

    Spatial and Spatio-Temporal Log-Gaussian Cox Processes: Extending the Geostatistical Paradigm , author=. Statistical Science , volume=. 2013 , publisher=

    2013

  11. [11]

    Journal of Statistical Software , volume=

    lgcp: an R package for inference with spatial and spatio-temporal log-Gaussian Cox processes , author=. Journal of Statistical Software , volume=

  12. [12]

    arXiv preprint arXiv:2104.03528 , year=

    Neural temporal point processes: A review , author=. arXiv preprint arXiv:2104.03528 , year=

    work page arXiv

  13. [13]

    arXiv preprint arXiv:2502.09341 , year=

    Neural spatiotemporal point processes: Trends and challenges , author=. arXiv preprint arXiv:2502.09341 , year=

    work page arXiv

  14. [14]

    Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining , pages=

    Recurrent marked temporal point processes: Embedding event history to vector , author=. Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining , pages=

  15. [16]

    Learning for Dynamics and Control Conference , pages=

    Neural point process for learning spatiotemporal event dynamics , author=. Learning for Dynamics and Control Conference , pages=. 2022 , organization=

    2022

  16. [17]

    Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining , pages=

    Deep mixture point processes: Spatio-temporal event prediction with rich contextual information , author=. Proceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining , pages=

  17. [18]

    Advances in Neural Information Processing Systems , volume=

    Integration-free training for spatio-temporal multimodal covariate deep kernel point processes , author=. Advances in Neural Information Processing Systems , volume=

  18. [20]

    Spatial Statistics , volume=

    Spatio-temporal point process statistics: a review , author=. Spatial Statistics , volume=. 2016 , publisher=

    2016

  19. [21]

    2023 , publisher=

    Spatial statistics for data science: theory and practice with R , author=. 2023 , publisher=

    2023

  20. [22]

    Scandinavian journal of statistics , volume=

    Log gaussian cox processes , author=. Scandinavian journal of statistics , volume=. 1998 , publisher=

    1998

  21. [24]

    Stochastic Environmental Research and Risk Assessment , volume=

    Spatio-temporal modeling of infectious diseases by integrating compartment and point process models , author=. Stochastic Environmental Research and Risk Assessment , volume=. 2023 , publisher=

    2023

  22. [25]

    Journal of Statistical Software , volume=

    Spatio-temporal analysis of epidemic phenomena using the R package surveillance , author=. Journal of Statistical Software , volume=

  23. [26]

    Journal of the American Statistical Association , volume=

    A spatio-temporal point process model for ambulance demand , author=. Journal of the American Statistical Association , volume=. 2015 , publisher=

    2015

  24. [27]

    The Annals of Applied Statistics , pages=

    PREDICTING MELBOURNE AMBULANCE DEMAND USING KERNEL WARPING , author=. The Annals of Applied Statistics , pages=. 2016 , publisher=

    2016

  25. [29]

    The Annals of Applied Statistics , volume=

    A nonseparable first-order spatiotemporal intensity for events on linear networks: An application to ambulance interventions , author=. The Annals of Applied Statistics , volume=. 2024 , publisher=

    2024

  26. [30]

    the 52nd Hawaii International Conference on System Sciences, 2019, January 8-11, Grand Wailea, Maui, Hawai'i , pages=

    Forecasting the Demand for Emergency Medical Services , author=. the 52nd Hawaii International Conference on System Sciences, 2019, January 8-11, Grand Wailea, Maui, Hawai'i , pages=. 2019 , organization=

    2019

  27. [31]

    IEEE Access , volume=

    Nonparametric method for modeling clustering phenomena in emergency calls under spatial-temporal self-exciting point processes , author=. IEEE Access , volume=. 2019 , publisher=

    2019

  28. [32]

    2003 , publisher=

    Statistical inference and simulation for spatial point processes , author=. 2003 , publisher=

    2003

  29. [33]

    2008 , publisher=

    Statistical analysis and modelling of spatial point patterns , author=. 2008 , publisher=

    2008

  30. [34]

    arXiv preprint arXiv:2207.07814 , year=

    Spatial point process via regularisation modelling of ambulance call risk , author=. arXiv preprint arXiv:2207.07814 , year=

    work page arXiv

  31. [35]

    Spatial Statistics , volume=

    Large-scale modelling and forecasting of ambulance calls in northern Sweden using spatio-temporal log-Gaussian Cox processes , author=. Spatial Statistics , volume=. 2020 , publisher=

    2020

  32. [36]

    Machine Learning , volume=

    Context-aware spatio-temporal event prediction via convolutional Hawkes processes , author=. Machine Learning , volume=. 2022 , publisher=

    2022

  33. [38]

    Journal of The Royal Society Interface , volume=

    Model building and assessment of the impact of covariates for disease prevalence mapping in low-resource settings: to explain and to predict , author=. Journal of The Royal Society Interface , volume=. 2021 , publisher=

    2021

  34. [39]

    Computers in Biology and Medicine , volume=

    A data-driven hybrid ensemble AI model for COVID-19 infection forecast using multiple neural networks and reinforced learning , author=. Computers in Biology and Medicine , volume=. 2022 , publisher=

    2022

  35. [40]

    Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , pages=

    Spatio-temporal diffusion point processes , author=. Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , pages=

  36. [42]

    Journal of Applied Statistics , pages=

    BSTPP: a python package for Bayesian spatiotemporal point processes , author=. Journal of Applied Statistics , pages=. 2025 , publisher=

    2025

  37. [43]

    Remote sensing of environment , volume=

    The Harmonized Landsat and Sentinel-2 surface reflectance data set , author=. Remote sensing of environment , volume=. 2018 , publisher=

    2018

  38. [44]

    Journal of the American statistical Association , volume=

    Strictly proper scoring rules, prediction, and estimation , author=. Journal of the American statistical Association , volume=. 2007 , publisher=

    2007

  39. [45]

    Statistics and computing , volume=

    Practical Bayesian model evaluation using leave-one-out cross-validation and WAIC , author=. Statistics and computing , volume=. 2017 , publisher=

    2017

  40. [46]

    Journal of the Royal Society Interface , volume=

    PriorVAE: encoding spatial priors with variational autoencoders for small-area estimation , author=. Journal of the Royal Society Interface , volume=. 2022 , publisher=

    2022

  41. [47]

    Spatio-temporal modeling of infectious diseases by integrating compartment and point process models

    Andr \'e Victor Ribeiro Amaral, Jonatan A Gonz \'a lez, and Paula Moraga. Spatio-temporal modeling of infectious diseases by integrating compartment and point process models. Stochastic Environmental Research and Risk Assessment , 37(4):1519--1533, 2023

    2023

  42. [48]

    AlphaEarth Foundations: An embedding field model for accurate and efficient global mapping from sparse label data

    Christopher F Brown, Michal R Kazmierski, Valerie J Pasquarella, William J Rucklidge, Masha Samsikova, Chenhui Zhang, Evan Shelhamer, Estefania Lahera, Olivia Wiles, Simon Ilyushchenko, et al. Alphaearth foundations: An embedding field model for accurate and efficient global mapping from sparse label data. arXiv preprint arXiv:2507.22291 , 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  43. [49]

    Neural spatio-temporal point processes

    Ricky TQ Chen, Brandon Amos, and Maximilian Nickel. Neural spatio-temporal point processes. arXiv preprint arXiv:2011.04583 , 2020

    work page arXiv 2011

  44. [50]

    The harmonized landsat and sentinel-2 surface reflectance data set

    Martin Claverie, Junchang Ju, Jeffrey G Masek, Jennifer L Dungan, Eric F Vermote, Jean-Claude Roger, Sergii V Skakun, and Christopher Justice. The harmonized landsat and sentinel-2 surface reflectance data set. Remote sensing of environment , 219:145--161, 2018

    2018

  45. [51]

    Spatial and spatio-temporal log-gaussian cox processes: Extending the geostatistical paradigm

    Peter J Diggle, Paula Moraga, Barry Rowlingson, and Benjamin M Taylor. Spatial and spatio-temporal log-gaussian cox processes: Extending the geostatistical paradigm. Statistical Science , 28(4), 2013

    2013

  46. [52]

    Statistical analysis of spatial and spatio-temporal point patterns

    Peter J Diggle. Statistical analysis of spatial and spatio-temporal point patterns . CRC press, 2013

    2013

  47. [53]

    A nonseparable first-order spatiotemporal intensity for events on linear networks: An application to ambulance interventions

    Andrea Gilardi, Riccardo Borgoni, and Jorge Mateu. A nonseparable first-order spatiotemporal intensity for events on linear networks: An application to ambulance interventions. The Annals of Applied Statistics , 18(1):529--554, 2024

    2024

  48. [54]

    Model building and assessment of the impact of covariates for disease prevalence mapping in low-resource settings: to explain and to predict

    Emanuele Giorgi, Claudio Fronterr \`e , Peter M Macharia, Victor A Alegana, Robert W Snow, and Peter J Diggle. Model building and assessment of the impact of covariates for disease prevalence mapping in low-resource settings: to explain and to predict. Journal of The Royal Society Interface , 18(179):20210104, 2021

    2021

  49. [55]

    Strictly proper scoring rules, prediction, and estimation

    Tilmann Gneiting and Adrian E Raftery. Strictly proper scoring rules, prediction, and estimation. Journal of the American statistical Association , 102(477):359--378, 2007

    2007

  50. [56]

    Spatio-temporal point process statistics: a review

    Jonatan A Gonz \'a lez, Francisco J Rodr \' guez-Cort \'e s, Ottmar Cronie, and Jorge Mateu. Spatio-temporal point process statistics: a review. Spatial Statistics , 18:505--544, 2016

    2016

  51. [57]

    Nonparametric method for modeling clustering phenomena in emergency calls under spatial-temporal self-exciting point processes

    Chenlong Li, Zhanjie Song, and Xu Wang. Nonparametric method for modeling clustering phenomena in emergency calls under spatial-temporal self-exciting point processes. IEEE Access , 7:24865--24876, 2019

    2019

  52. [58]

    u dke, Enric Rabasseda Ravent \'o s, Marcel Kollovieh, and Stephan G \

    David L \"u dke, Enric Rabasseda Ravent \'o s, Marcel Kollovieh, and Stephan G \"u nnemann. Unlocking point processes through point set diffusion. arXiv preprint arXiv:2410.22493 , 2024

    work page arXiv 2024

  53. [59]

    Bstpp: a python package for bayesian spatiotemporal point processes

    Isaac Manring, Honglang Wang, George Mohler, and Xenia Miscouridou. Bstpp: a python package for bayesian spatiotemporal point processes. Journal of Applied Statistics , pages 1--20, 2025

    2025

  54. [60]

    Transfeat-tpp: An interpretable deep covariate temporal point processes

    Zizhuo Meng, Boyu Li, Xuhui Fan, Zhidong Li, Yang Wang, Fang Chen, and Feng Zhou. Transfeat-tpp: An interpretable deep covariate temporal point processes. arXiv preprint arXiv:2407.16161 , 2024

    work page arXiv 2024

  55. [61]

    Spatio-temporal analysis of epidemic phenomena using the r package surveillance

    Sebastian Meyer, Leonhard Held, and Michael H \"o hle. Spatio-temporal analysis of epidemic phenomena using the r package surveillance. Journal of Statistical Software , 77:1--55, 2017

    2017

  56. [62]

    Log gaussian cox processes

    Jesper M ller, Anne Randi Syversveen, and Rasmus Plenge Waagepetersen. Log gaussian cox processes. Scandinavian journal of statistics , 25(3):451--482, 1998

    1998

  57. [63]

    Spatial statistics for data science: theory and practice with R

    Paula Moraga. Spatial statistics for data science: theory and practice with R . Chapman and Hall/CRC, 2023

    2023

  58. [64]

    A review of self-exciting spatio-temporal point processes and their applications

    Alex Reinhart. A review of self-exciting spatio-temporal point processes and their applications. Statistical Science , 33(3):299--318, 2018

    2018

  59. [65]

    Priorvae: encoding spatial priors with variational autoencoders for small-area estimation

    Elizaveta Semenova, Yidan Xu, Adam Howes, Theo Rashid, Samir Bhatt, Swapnil Mishra, and Seth Flaxman. Priorvae: encoding spatial priors with variational autoencoders for small-area estimation. Journal of the Royal Society Interface , 19(191):20220094, 2022

    2022

  60. [66]

    Forecasting the demand for emergency medical services

    Krisjanis Steins, Niki Matinrad, and Tobias Andersson Granberg. Forecasting the demand for emergency medical services. In the 52nd Hawaii International Conference on System Sciences, 2019, January 8-11, Grand Wailea, Maui, Hawai'i , pages 1855--1864. University of Hawai'i at Manoa, 2019

    2019

  61. [67]

    Practical bayesian model evaluation using leave-one-out cross-validation and waic

    Aki Vehtari, Andrew Gelman, and Jonah Gabry. Practical bayesian model evaluation using leave-one-out cross-validation and waic. Statistics and computing , 27(5):1413--1432, 2017

    2017

  62. [68]

    Twenty ways to estimate the log gaussian cox process model with point and aggregated case data: the rts2 package for r

    Samuel I Watson. Twenty ways to estimate the log gaussian cox process model with point and aggregated case data: the rts2 package for r. arXiv preprint arXiv:2403.09448 , 2024

    work page arXiv 2024

  63. [69]

    Spatio-temporal diffusion point processes

    Yuan Yuan, Jingtao Ding, Chenyang Shao, Depeng Jin, and Yong Li. Spatio-temporal diffusion point processes. In Proceedings of the 29th ACM SIGKDD Conference on Knowledge Discovery and Data Mining , pages 3173--3184, 2023

    2023

  64. [70]

    Predicting ambulance demand: A spatio-temporal kernel approach

    Zhengyi Zhou and David S Matteson. Predicting ambulance demand: A spatio-temporal kernel approach. In Proceedings of the 21th ACM SIGKDD international conference on knowledge discovery and data mining , pages 2297--2303, 2015

    2015

  65. [71]

    Predicting melbourne ambulance demand using kernel warping

    Zhengyi Zhou and David S Matteson. Predicting melbourne ambulance demand using kernel warping. The Annals of Applied Statistics , pages 1977--1996, 2016

    1977

  66. [72]

    A spatio-temporal point process model for ambulance demand

    Zhengyi Zhou, David S Matteson, Dawn B Woodard, Shane G Henderson, and Athanasios C Micheas. A spatio-temporal point process model for ambulance demand. Journal of the American Statistical Association , 110(509):6--15, 2015

    2015

  67. [73]

    Predicting Ambulance Demand: Challenges and Methods

    Zhengyi Zhou. Predicting ambulance demand: challenges and methods. arXiv preprint arXiv:1606.05363 , 2016

    work page internal anchor Pith review Pith/arXiv arXiv 2016

  68. [74]

    Structure and Interpretation of Computer Programs

    Harold Abelson and Gerald Jay Sussman and Julie Sussman. Structure and Interpretation of Computer Programs. 1985

    1985

  69. [75]

    Visual Information Extraction with Lixto

    Robert Baumgartner and Georg Gottlob and Sergio Flesca. Visual Information Extraction with Lixto. Proceedings of the 27th International Conference on Very Large Databases. 2001

    2001

  70. [76]

    Brachman and James G

    Ronald J. Brachman and James G. Schmolze. An overview of the KL-ONE knowledge representation system. Cognitive Science. 1985

    1985

  71. [77]

    Complexity results for nonmonotonic logics

    Georg Gottlob. Complexity results for nonmonotonic logics. Journal of Logic and Computation. 1992

    1992

  72. [78]

    Hypertree Decompositions and Tractable Queries

    Georg Gottlob and Nicola Leone and Francesco Scarcello. Hypertree Decompositions and Tractable Queries. Journal of Computer and System Sciences. 2002

    2002

  73. [79]

    Levesque

    Hector J. Levesque. Foundations of a functional approach to knowledge representation. Artificial Intelligence. 1984

    1984

  74. [80]

    Levesque

    Hector J. Levesque. A logic of implicit and explicit belief. Proceedings of the Fourth National Conference on Artificial Intelligence. 1984

    1984

  75. [81]

    On the compilability and expressive power of propositional planning formalisms

    Bernhard Nebel. On the compilability and expressive power of propositional planning formalisms. Journal of Artificial Intelligence Research. 2000

    2000