Daniel Jay Alford-Lago

Dr. Daniel Jay Alford-Lago

orcid

Research Focus & Metrics

Academic Career Length

6Years (2021-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
1 March 2022 in Chaos: An Interdisciplinary Journal of Nonlinear Science
Verified
Journal Article
Open access
PDF file
DADaniel Jay Alford-Lago
Daniel Jay Alford-Lago
CCC. W. Curtis
C. W. Curtis
AIA. T. Ihler
A. T. Ihler
OIO. Issan
O. Issan
Koopman operator theory shows how nonlinear dynamical systems can be represented as an infinite-dimensional, linear operator acting on a Hilbert space of observables of the system. However, determining the relevant modes and eigenvalues of this infinite-dimensional operator can be difficult. The extended dynamic mode decomposition (EDMD) is one such method for generating approximations to Koopman spectra and modes, but the EDMD method faces its own set of challenges due to the need of user defined observables. To address this issue, we explore the use of autoencoder networks to simultaneously find optimal families of observables, which also generate both accurate embeddings of the flow into a space of observables and submersions of the observables back into flow coordinates. This network results in a global transformation of the flow and affords future state prediction via the EDMD and the decoder network. We call this method the deep learning dynamic mode decomposition (DLDMD). The method is tested on canonical nonlinear data sets and is shown to produce results that outperform a standard DMD approach and enable data-driven prediction where the standard DMD fails.

4 January 2021 in Physical Review E
Verified
Journal Article
Closed access
PDF file
Christopher CurtisChristopher Curtis
Christopher Curtis
DADaniel Jay Alford-Lago
Daniel Jay Alford-Lago
Dynamic-mode decomposition (DMD) is a well-established data-driven method of finding temporally evolving linear-mode decompositions of a nonlinear time series. Traditionally, this method presumes that all relevant dimensions are sampled through measurement. To address dynamical systems in which the data may be incomplete or represent only partial observation of a more complex system, we extend the DMD algorithm by including a Mori-Zwanzig decomposition to derive memory kernels that capture the averaged dynamics of the unresolved variables as projected onto the resolved dimensions. From this, we then derive what we call the memory-dependent dynamic mode decomposition (MDDMD). Through numerical examples, the MDDMD method is shown to produce reasonable approximations of the ensemble-averaged dynamics of the full system given a single time series measurement of the resolved variables.

1 August 2023 in Chaos: An Interdisciplinary Journal of Nonlinear Science
Verified
Journal Article
Open access
PDF file
Christopher CurtisChristopher Curtis
Christopher Curtis
DADaniel Jay Alford-Lago
Daniel Jay Alford-Lago
EBErik Bollt
Erik Bollt
ATAndrew Tuma
Andrew Tuma
While the acquisition of time series has become more straightforward, developing dynamical models from time series is still a challenging and evolving problem domain. Within the last several years, to address this problem, there has been a merging of machine learning tools with what is called the dynamic-mode decomposition (DMD). This general approach has been shown to be an especially promising avenue for accurate model development. Building on this prior body of work, we develop a deep learning DMD based method, which makes use of the fundamental insight of Takens’ embedding theorem to build an adaptive learning scheme that better approximates higher dimensional and chaotic dynamics. We call this method the Deep Learning Hankel DMD. We likewise explore how our method learns mappings, which tend, after successful training, to significantly change the mutual information between dimensions in the dynamics. This appears to be a key feature in enhancing DMD overall, and it should help provide further insight into developing other deep learning methods for time series analysis and model generation.

18 September 2024 in International Journal of Bifurcation and Chaos
Verified
Journal Article
Closed access
Christopher CurtisChristopher Curtis
Christopher Curtis
EBErik Bollt
Erik Bollt
DADaniel Jay Alford-Lago
Daniel Jay Alford-Lago

3 August 2023 in Radio Science
Verified
Journal Article
Open access
DADaniel Jay Alford-Lago
Daniel Jay Alford-Lago
CCChristopher W. Curtis
Christopher W. Curtis
AIAlexander T. Ihler
Alexander T. Ihler
KZKatherine A. Zawdie
Katherine A. Zawdie
We present a method for forecasting the foF2 and hmF2 parameters using modal decompositions from measured ionospheric electron density profiles. Our method is based on Dynamic Mode Decomposition (DMD), which provides a means of determining spatiotemporal modes from measurements alone. Our proposed extensions to DMD use wavelet decompositions that provide separation of a wide range of high‐intensity, transient temporal scales in the measured data. This scale separation allows for DMD models to be fit on each scale individually, and we show that together they generate a more accurate forecast of the time‐evolution of the F‐layer peak. We call this method the Scale‐Separated Dynamic Mode Decomposition (SSDMD). The approach is shown to produce stable modes that can be used as a time‐stepping model to predict the state of foF2 and hmF2 at a high time resolution. We demonstrate the SSDMD method on data sets covering periods of high and low solar activity as well as low, mid, and high latitude locations.

6 January 2026 in Journal of Geophysical Research: Machine Learning and Computation
Verified
Journal Article
Open access
DADaniel Jay Alford-Lago
Daniel Jay Alford-Lago
CCC. W. Curtis
C. W. Curtis
AIA. T. Ihler
A. T. Ihler
DDD. P. Drob
D. P. Drob
KZK. A. Zawdie
K. A. Zawdie
We present a novel method for forecasting key ionospheric parameters using transformer‐based neural networks. The model provides accurate forecasts and uncertainty quantification of the F2‐layer peak plasma frequency (foF2), the F2‐layer peak density height (hmF2), and total electron content for a given geographic location. It supports a number of exogenous variables, including F10.7 cm solar flux and disturbance storm time (Dst). We demonstrate how transformers can be trained in a data assimilation‐like fashion that uses these exogenous variables along with naïve predictions from climatology to generate 24‐hr forecasts with nonparametric uncertainty bounds. We call this method the Local Ionospheric Forecast Transformer. We demonstrate that the trained model can generalize to new geographic locations and time periods not seen during training, and we compare its performance to that of the International Reference Ionosphere using CCIR coefficients.