Paper: RoughPy

papers/sam_morley/myst.yml

@@ -0,0 +1,74 @@
+# See docs at: https://mystmd.org/guide/frontmatter
+version: 1
+project:
+  id: scipy-2024-sam_morley
+  title: RoughPy
+  subtitle: Streaming data is rarely smooth
+  short_title: RoughPy
+
+  authors: 
+    - name: Sam Morley
+      email: [email protected]
+      orcid: 0000-0001-5971-7418
+      affiliations: 
+        - University Of Oxford
+      roles:
+        - Conceptualization
+        - Writing - original draft
+        - Software
+      corresponding: true
+    - name: Terry Lyons
+      email: [email protected]
+      orcid: 0000-0002-9972-2809
+      affiliations:
+        - University of Oxford
+      roles:
+        - Conceptualization
+        - Supervision
+        - Writing - review & editing
+
+  # description:
+  keywords: 
+    - sequential data
+    - unparametrised paths
+    - time series 
+    - rough paths
+    - signatures 
+    - data science 
+    - machine learning
+    - signature kernels
+    - Log-ODE method
+  github: datasig-ac-uk/roughpy
+  bibliography: 
+    - roughpy.bib
+  # Add the abbreviations that you use in your paper here
+  abbreviations:
+    RFI: Radio frequency interference
+    NLP: Natural Language Processing
+    ODE: Ordinary Differential Equation
+    CDE: Controlled differential equation
+    PDE: Partial differential equation
+
+  # It is possible to explicitly ignore the `doi-exists` check for certain citation keys
+  error_rules:
+    - rule: doi-exists
+      severity: ignore
+      keys:
+        - Lyons1998
+        - esig
+        - dlpack
+        - pybind11
+        - coropa_project
+        - Granlund12
+        - NEURIPS2021_18a9042b
+
+  # A banner will be generated for you on publication, this is a placeholder
+  banner: banner.png
+  # The rest of the information shouldn't be modified
+  subject: Research Article
+  open_access: true
+  license: CC-BY-4.0
+  venue: Scipy 2024
+  date: 2024-07-10
+site:
+  template: article-theme

papers/sam_morley/roughpy.bib

@@ -0,0 +1,140 @@
+@inproceedings{tseriotou_etal_2024_sig,
+    title = "Sig-Networks Toolkit: Signature Networks for Longitudinal Language Modelling",
+    author = "Tseriotou, Talia  and
+      Chan, Ryan  and
+      Tsakalidis, Adam  and
+      Bilal, Iman Munire  and
+      Kochkina, Elena  and
+      Lyons, Terry  and
+      Liakata, Maria",
+    editor = "Aletras, Nikolaos  and
+      De Clercq, Orphee",
+    booktitle = "Proceedings of the 18th Conference of the European Chapter of the Association for Computational Linguistics: System Demonstrations",
+    month = mar,
+    year = "2024",
+    address = "St. Julians, Malta",
+    publisher = "Association for Computational Linguistics",
+    url = "https://aclanthology.org/2024.eacl-demo.24",
+    pages = "223--237",
+    abstract = "We present an open-source, pip installable toolkit, Sig-Networks, the first of its kind for longitudinal language modelling. A central focus is the incorporation of Signature-based Neural Network models, which have recently shown success in temporal tasks. We apply and extend published research providing a full suite of signature-based models. Their components can be used as PyTorch building blocks in future architectures. Sig-Networks enables task-agnostic dataset plug-in, seamless preprocessing for sequential data, parameter flexibility, automated tuning across a range of models. We examine signature networks under three different NLP tasks of varying temporal granularity: counselling conversations, rumour stance switch and mood changes in social media threads, showing SOTA performance in all three, and provide guidance for future tasks. We release the Toolkit as a PyTorch package with an introductory video, Git repositories for preprocessing and modelling including sample notebooks on the modeled NLP tasks.",
+    doi = "10.48550/arXiv.2312.03523",
+}
+
+@inproceedings{NEURIPS2020_4a5876b4,
+ author = {Kidger, Patrick and Morrill, James and Foster, James and Lyons, Terry},
+ booktitle = {Advances in Neural Information Processing Systems},
+ editor = {H. Larochelle and M. Ranzato and R. Hadsell and M.F. Balcan and H. Lin},
+ pages = {6696--6707},
+ publisher = {Curran Associates, Inc.},
+ title = {Neural Controlled Differential Equations for Irregular Time Series},
+ url = {https://proceedings.neurips.cc/paper_files/paper/2020/file/4a5876b450b45371f6cfe5047ac8cd45-Paper.pdf},
+ volume = {33},
+ year = {2020},
+ doi = "10.48550/arXiv.1906.08215",
+}
+
+@article{Lyons1998,
+abstract = {This paper aims to provide a systematic approach to the treatment of differential equations of the typedyt = Σi fi(yt) dxti where the driving signal xt is a rough path. Such equations are very common and occur particularly frequently in probability where the driving signal might be a vector valued Brownian motion, semi-martingale or similar process.However, our approach is deterministic, is totally independent of probability and permits much rougher paths than the Brownian paths usually discussed. The results here are strong enough to treat the main probabilistic examples and significantly widen the class of stochastic processes which can be used to drive stochastic differential equations. (For a simple example see [10], [1]).We hope our results will have an influence on infinite dimensional analysis on path spaces, loop groups, etc. as well as in more applied situations. Variable step size algorithms for the numerical integration of stochastic differential equations [8] have been constructed as a consequence of these results.},
+author = {Lyons, Terry J.},
+journal = {Revista Matemática Iberoamericana},
+keywords = {Ecuaciones diferenciales estocásticas; Proceso de difusión; Movimiento browniano; stochastic differential equations; rough paths; Brown motion; Gauss and Markov processes; Lie algebras},
+language = {eng},
+number = {2},
+pages = {215-310},
+title = {Differential equations driven by rough signals.},
+url = {http://eudml.org/doc/39555},
+volume = {14},
+year = {1998},
+}
+
+@inproceedings{liao2021a,
+  edition = {},
+  number = {},
+  journal = {},
+  pages = {},
+  publisher = {British Machine Vision Association},
+  school = {},
+  title = {Logsig-RNN: a novel network for robust and efficient skeleton-based action recognition},
+  volume = {},
+  author = {Liao, S and Lyons, TJ and Yang, W and Schlegel, K and Ni, H},
+  editor = {},
+  year = {2021},
+  organizer = {32nd British Machine Vision Conference (BMVC 2021)},
+ series = {},
+ doi = "10.48550/arXiv.2110.13008",
+}
+
+@inproceedings{NEURIPS2021_18a9042b,
+ author = {Fermanian, Adeline and Marion, Pierre and Vert, Jean-Philippe and Biau, G\'{e}rard},
+ booktitle = {Advances in Neural Information Processing Systems},
+ editor = {M. Ranzato and A. Beygelzimer and Y. Dauphin and P.S. Liang and J. Wortman Vaughan},
+ pages = {3121--3134},
+ publisher = {Curran Associates, Inc.},
+ title = {Framing RNN as a kernel method: A neural ODE approach},
+ url = {https://proceedings.neurips.cc/paper_files/paper/2021/file/18a9042b3fc5b02fe3d57fea87d6992f-Paper.pdf},
+ volume = {34},
+ year = {2021},
+}
+
+@article{JMLR_v20_16_314,
+  author  = {Franz J. Kiraly and Harald Oberhauser},
+  title   = {Kernels for Sequentially Ordered Data},
+  journal = {Journal of Machine Learning Research},
+  year    = {2019},
+  volume  = {20},
+  number  = {31},
+  pages   = {1--45},
+  url     = {http://jmlr.org/papers/v20/16-314.html},
+  doi = "10.48550/arXiv.2102.03657",
+}
+
+@misc{pybind11,
+   author = {Wenzel Jakob and Jason Rhinelander and Dean Moldovan},
+   year = {2017},
+   note = {https://github.com/pybind/pybind11},
+   title = {pybind11 -- Seamless operability between C++11 and Python}
+}
+
+@misc{esig,
+  author = {Terry Lyons and David Maxwell},
+  year = 2017,
+  note = {https://github.com/datasig-ac-uk/esig},
+  title = {esig}
+}
+
+@misc{coropa_project,
+  author = {Stephen Buckley and Djalil Chafai and Greg Gyurk\`{o} and Arend
+    Janssen and Christophe Ladroue and Christian Litterer and Terry Lyons 
+    and ChangLiang Xu},
+  year = 2006,
+  note = {https://coropa.sourceforge.io},
+  title = {CoRoPa -- Computational Rough Paths project}
+}
+
+@inproceedings{NEURIPS2019_d2cdf047,
+ author = {Kidger, Patrick and Bonnier, Patric and Perez Arribas, Imanol and Salvi, Cristopher and Lyons, Terry},
+ booktitle = {Advances in Neural Information Processing Systems},
+ editor = {H. Wallach and H. Larochelle and A. Beygelzimer and F. d\textquotesingle Alch\'{e}-Buc and E. Fox and R. Garnett},
+ pages = {},
+ publisher = {Curran Associates, Inc.},
+ title = {Deep Signature Transforms},
+ url = {https://proceedings.neurips.cc/paper_files/paper/2019/file/d2cdf047a6674cef251d56544a3cf029-Paper.pdf},
+ volume = {32},
+ year = {2019},
+ doi = "10.48550/arXiv.1905.08494"
+}
+
+@misc{dlpack, 
+  author = {DLPack},
+  year = 2023,
+  title = {Open In Memory Tensor structure},
+  note = {https://dmlc.github.io/dlpack/latest/}
+}
+
+@Manual{Granlund12,
+  title = "{GNU MP}: {T}he {GNU} {M}ultiple {P}recision {A}rithmetic {L}ibrary",
+  author = "Torbjörn Granlund and {the GMP development team}",
+  edition = "5.0.5",
+  year = 	 2012,
+  note = "http://gmplib.org/"
+}