arXiv subjects

Michael Casey

Publications and source records attributed to Michael Casey.

8 recordsLinked to original sources

arXiv2026

Tulip-Shaped Orbits for Lunar South-Pole PNT and Direct-to-Earth Relay Missions

This geometric study evaluates a compact seven-petal, 6:5-resonant tulip-shaped orbit constellation for lunar south-pole positioning, navigation, and timing (PNT) and direct-to-Earth relay services. The tulip-shaped orbits are compared against elliptical lunar frozen orbit (ELFO) constellations over the NASA LunaNet Service Volume 2 (SV2), covering lunar latitudes south of -75 degrees. We compare a six-satellite tulip baseline with a minimum-cost five-satellite variant; both use the same shared three-body orbit and differ only in satellite count and along-track phasing. Performance is scored against three Initial Operating Capability C (IOC-C) metrics: line-of-sight (LOS) link availability, Lunar Augmented Navigation System (LANS) geometric dilution of precision (GDOP) below 6, and daily extravehicular activity (EVA) usable-PNT windows. Both tulip constellations satisfy all three IOC-C metrics across SV2. The six-satellite configuration meets requirements with wide margin: 75% worst-point daily GDOP availability and 18 hours of daily EVA support. The five-satellite variant also passes, but with thinner margin: 44% availability and 10 hours of EVA support. Unlike the ELFO configurations, each spacecraft in the tulip-shaped constellation maintains continuous Earth line of sight, providing persistent geometric opportunity for direct single-hop Earth relay. Because all spacecraft share a single three-body orbit, initial phasing and post-failure reconstitution reduce to along-track drift maneuvers between neighboring orbits, with screening-level estimates indicating low maneuver cost.

eess.SY↗

arXiv2025

A Graph Engine for Guitar Chord-Tone Soloing Education

We present a graph-based engine for computing chord tone soloing suggestions for guitar students. Chord tone soloing is a fundamental practice for improvising over a chord progression, where the instrumentalist uses only the notes contained in the current chord. This practice is a building block for all advanced jazz guitar theory but is difficult to learn and practice. First, we discuss methods for generating chord-tone arpeggios. Next, we construct a weighted graph where each node represents a chord tone arpeggio for a chord in the progression. Then, we calculate the edge weight between each consecutive chord's nodes in terms of optimal transition tones. We then find the shortest path through this graph and reconstruct a chord-tone soloing line. Finally, we discuss a user-friendly system to handle input and output to this engine for guitar students to practice chord tone soloing.

cs.AI↗

arXiv2025

Topological Data Analysis for Unsupervised Feature Selection in Large Scale Spatial Omics Data Sets

Spatial transcriptomics studies are becoming increasingly large and commonplace, necessitating simultaneous analysis of a large number of spatially resolved variables. Correspondingly, a diverse range of methodologies have been proposed to compare the spatial expression structure of genes. Here, we apply persistent homology, a method from topological data analysis, to produce a continuous quantification of spatial structure in a given gene's expression, and show how this can be used for downstream tasks such as spatially variable gene identification. We explore the unique advantages of topology for this task, deriving biologically meaningful insights into kidney disease and myocardial infarction using public spatial transcriptomics data. We also show how the non-parametric nature of homology enables our methodology to extend naturally to other spatial omics modalities, demonstrating this on a spatial metabolomics sample. Our work showcases the advantages of using a continuous quantification of spatial structure over p-value based approaches to SVG identification, the potential for developing unified methods for the analysis of different spatial omics modalities, and the utility of persistent homology in big data applications.

q-bio.QM↗

arXiv2025

Notes on Quantum Soundscapes and Music

We describe our investigations concerning the sonification of measured data from experiments involving various mesoscopic mechanical oscillator systems cooled down close to their quantum ground states, and music generation from a programmed quantum computer that subjects a single quantum bit ("qubit") to various unitary rotations, composed in order to test for the breakdown of macroscopic realism as expressed by the violation of the Leggett-Garg inequality. "Listening'' to data via their resulting sonifications facilitates the discovery of signals that might otherwise be hard to detect in common graphic (i.e., visual) representations, and for the quantum computer music experiment provides a complementary way to discern when the measured qubit data violates macroscopic realism with some prior listening training. The resulting soundscapes and music also provide a complementary window into the quantum realm that is accessible to non-experts with open ears.

quant-ph↗

arXiv2023

Sequential Transfer Learning to Decode Heard and Imagined Timbre from fMRI Data

We present a sequential transfer learning framework for transformers on functional Magnetic Resonance Imaging (fMRI) data and demonstrate its significant benefits for decoding musical timbre. In the first of two phases, we pre-train our stacked-encoder transformer architecture on Next Thought Prediction, a self-supervised task of predicting whether or not one sequence of fMRI data follows another. This phase imparts a general understanding of the temporal and spatial dynamics of neural activity, and can be applied to any fMRI dataset. In the second phase, we fine-tune the pre-trained models and train additional fresh models on the supervised task of predicting whether or not two sequences of fMRI data were recorded while listening to the same musical timbre. The fine-tuned models achieve significantly higher accuracy with shorter training times than the fresh models, demonstrating the efficacy of our framework for facilitating transfer learning on fMRI data. Additionally, our fine-tuning task achieves a level of classification granularity beyond standard methods. This work contributes to the growing literature on transformer architectures for sequential transfer learning on fMRI data, and provides evidence that our framework is an improvement over current methods for decoding timbre.

q-bio.QM↗

arXiv2023

Decoding Imagined Auditory Pitch Phenomena with an Autoencoder Based Temporal Convolutional Architecture

Stimulus decoding of functional Magnetic Resonance Imaging (fMRI) data with machine learning models has provided new insights about neural representational spaces and task-related dynamics. However, the scarcity of labelled (task-related) fMRI data is a persistent obstacle, resulting in model-underfitting and poor generalization. In this work, we mitigated data poverty by extending a recent pattern-encoding strategy from the visual memory domain to our own domain of auditory pitch tasks, which to our knowledge had not been done. Specifically, extracting preliminary information about participants' neural activation dynamics from the unlabelled fMRI data resulted in improved downstream classifier performance when decoding heard and imagined pitch. Our results demonstrate the benefits of leveraging unlabelled fMRI data against data poverty for decoding pitch based tasks, and yields novel significant evidence for both separate and overlapping pathways of heard and imagined pitch processing, deepening our understanding of auditory cognitive neuroscience.

q-bio.NC↗

arXiv2023

Self-Supervised Pretraining on Paired Sequences of fMRI Data for Transfer Learning to Brain Decoding Tasks

In this work we introduce a self-supervised pretraining framework for transformers on functional Magnetic Resonance Imaging (fMRI) data. First, we pretrain our architecture on two self-supervised tasks simultaneously to teach the model a general understanding of the temporal and spatial dynamics of human auditory cortex during music listening. Our pretraining results are the first to suggest a synergistic effect of multitask training on fMRI data. Second, we finetune the pretrained models and train additional fresh models on a supervised fMRI classification task. We observe significantly improved accuracy on held-out runs with the finetuned models, which demonstrates the ability of our pretraining tasks to facilitate transfer learning. This work contributes to the growing body of literature on transformer architectures for pretraining and transfer learning with fMRI data, and serves as a proof of concept for our pretraining tasks and multitask pretraining on fMRI data.

cs.LG↗

arXiv2016

Log-Concave Duality in Estimation and Control

In this paper we generalize the estimation-control duality that exists in the linear-quadratic-Gaussian setting. We extend this duality to maximum a posteriori estimation of the system's state, where the measurement and dynamical system noise are independent log-concave random variables. More generally, we show that a problem which induces a convex penalty on noise terms will have a dual control problem. We provide conditions for strong duality to hold, and then prove relaxed conditions for the piecewise linear-quadratic case. The results have applications in estimation problems with nonsmooth densities, such as log-concave maximum likelihood densities. We conclude with an example reconstructing optimal estimates from solutions to the dual control problem, which has implications for sharing solution methods between the two types of problems.

math.OC↗