-
Stability and performance guarantees for misspecified multivariate score-driven filters
Authors:
Simon Donker van Heel,
Rutger-Jan Lange,
Dick van Dijk,
Bram van Os
Abstract:
We address the problem of tracking multivariate unobserved time-varying parameters under potential model misspecification. Specifically, we examine implicit and explicit score-driven (ISD and ESD) filters, which update parameter predictions using the gradient of the postulated logarithmic observation density (commonly referred to as the score). For both filter types, we derive novel sufficient con…
▽ More
We address the problem of tracking multivariate unobserved time-varying parameters under potential model misspecification. Specifically, we examine implicit and explicit score-driven (ISD and ESD) filters, which update parameter predictions using the gradient of the postulated logarithmic observation density (commonly referred to as the score). For both filter types, we derive novel sufficient conditions that ensure the invertibility of the filtered parameter path and the existence of a finite mean squared error (MSE) bound relative to the pseudo-true parameter path. Our (non-)asymptotic MSE bounds rely on mild moment conditions on the data-generating process, while our invertibility result is agnostic about the true process. For the ISD filter, concavity of the postulated log density combined with simple parameter restrictions is sufficient (though not necessary) to guarantee stability. In contrast, the ESD filter additionally requires the score to be Lipschitz continuous. We validate our theoretical findings and highlight the superior stability and performance of ISD over ESD filters through extensive simulation studies. Finally, we demonstrate the practical relevance of our approach through an empirical application to U.S. Treasury-bill rates.
△ Less
Submitted 18 April, 2025; v1 submitted 7 February, 2025;
originally announced February 2025.
-
Embedding digital chronotherapy into medical devices -- A canine validation for controlling status epilepticus through multi-scale rhythmic brain stimulation
Authors:
Mayela Zamora,
Sebastian Meller,
Filip Kajin,
James J Sermon,
Robert Toth,
Moaad Benjaber,
Derk-Jan Dijk,
Rafal Bogacz,
Gregory A Worrell,
Antonio Valentin,
Benoit Duchet,
Holger A Volk,
Timothy Denison
Abstract:
Circadian and other physiological rhythms play a key role in both normal homeostasis and disease processes. Such is the case of circadian and infradian seizure patterns observed in epilepsy. In this paper we explore a new implantable stimulator that implements chronotherapy as a feedforward input to supplement both open-loop and closed-loop methods. This integrated algorithm allows for stimulation…
▽ More
Circadian and other physiological rhythms play a key role in both normal homeostasis and disease processes. Such is the case of circadian and infradian seizure patterns observed in epilepsy. In this paper we explore a new implantable stimulator that implements chronotherapy as a feedforward input to supplement both open-loop and closed-loop methods. This integrated algorithm allows for stimulation to be adjusted to the ultradian, circadian and infradian patterns observed in patients through slowly-varying temporal adjustments of stimulation and algorithm sub-components, while also enabling adaption of stimulation based on immediate physiological needs such as a breakthrough seizure or change of posture. Embedded physiological sensors in the stimulator can be used to refine the baseline stimulation circadian pattern as a "digital zeitgeber". This approach is tested on a canine with severe drug-resistant idiopathic generalized epilepsy exhibiting a diurnal pattern correlated with sleep-wake cycles. Prior to implantation, the canine's cluster seizures evolved to status epilepticus (SE) and required emergency pharmacological intervention. The cranially-mounted system was fully-implanted bilaterally into the centromedian nucleus of the thalamus. Using time-based modulation, thalamocortical rhythm-specific tuning of frequency parameters as well as fast-adaptive modes based on activity, the canine experienced no further SE events post-implant as of the time of writing (seven months). Importantly, no significant cluster seizures have been observed either, allowing the reduction of rescue medication. The use of chronotherapy as a feedforward signal to augment adaptive neurostimulators could prove a useful method in conditions where sensitivity to temporal patterns are characteristics of the disease state, providing a novel mechanism for tailoring a more patient-specific therapy approach.
△ Less
Submitted 14 September, 2021; v1 submitted 6 July, 2021;
originally announced July 2021.
-
Self-supervised edge features for improved Graph Neural Network training
Authors:
Arijit Sehanobish,
Neal G. Ravindra,
David van Dijk
Abstract:
Graph Neural Networks (GNN) have been extensively used to extract meaningful representations from graph structured data and to perform predictive tasks such as node classification and link prediction. In recent years, there has been a lot of work incorporating edge features along with node features for prediction tasks. One of the main difficulties in using edge features is that they are often han…
▽ More
Graph Neural Networks (GNN) have been extensively used to extract meaningful representations from graph structured data and to perform predictive tasks such as node classification and link prediction. In recent years, there has been a lot of work incorporating edge features along with node features for prediction tasks. One of the main difficulties in using edge features is that they are often handcrafted, hard to get, specific to a particular domain, and may contain redundant information. In this work, we present a framework for creating new edge features, applicable to any domain, via a combination of self-supervised and unsupervised learning. In addition to this, we use Forman-Ricci curvature as an additional edge feature to encapsulate the local geometry of the graph. We then encode our edge features via a Set Transformer and combine them with node features extracted from popular GNN architectures for node classification in an end-to-end training scheme. We validate our work on three biological datasets comprising of single-cell RNA sequencing data of neurological disease, \textit{in vitro} SARS-CoV-2 infection, and human COVID-19 patients. We demonstrate that our method achieves better performance on node classification tasks over baseline Graph Attention Network (GAT) and Graph Convolutional Network (GCN) models. Furthermore, given the attention mechanism on edge and node features, we are able to interpret the cell types and genes that determine the course and severity of COVID-19, contributing to a growing list of potential disease biomarkers and therapeutic targets.
△ Less
Submitted 23 June, 2020;
originally announced July 2020.
