Search

Training neural network classifiers on datasets contaminated with noisy labels significantly increases the risk of overfitting. Thus, effectively implementing Early Stopping in noisy label environments is crucial. Under ideal circumstances, Early Stopping utilises a validation set uncorrupted by label noise to effectively monitor generalisation during training. However, obtaining a noise-free validation dataset can be costly and challenging to obtain. This study establishes that, in many typical learning environments, a noise-free validation set is not necessary for effective Early Stopping. Instead, near-optimal results can be achieved by monitoring accuracy on a noisy dataset - drawn from the same distribution as the noisy training set. Referred to as `Noisy Early Stopping' (NES), this method simplifies and reduces the cost of implementing Early Stopping. We provide theoretical insights into the conditions under which this method is effective and empirically demonstrate its robust performance across standard benchmarks using common loss functions.

This paper describes in detail how erasure codes are implemented in the Swarm system. First, in Section 1, we introduce erasure codes, and show how to apply them to files in Swarm (Section 2). In Section 3, we introduce security levels of data availability and derive their respective parameterisations. In Section 4, we describe a construct that enables cross-neighbourhood redundancy for singleton chunks and which completes erasure coding. Finally, in 5, we propose a number of retrieval strategies applicable to erasure-coded files., Comment: 14 pages, 7 figures, 4 tables

The identification of cells and particles based on their transport properties in microfluidic devices is crucial for numerous applications in biology and medicine. Neutrally buoyant particles transported in microfluidic channels, migrate laterally towards stable locations. However, the effect of the particle and flow properties on these focusing positions remain largely unknown. We conduct large scale simulations with dissipative particle dynamics, demonstrating that freely moving particles exhibit significant differences in their focusing patterns from particles that are prevented from rotation. In circular pipes, we observe drastic changes in rotating versus non-rotating focusing positions. We demonstrate that rotation-induced lateral lift force is significant, unlike previously believed, and is linearly dependent on the rotation magnitude. A simple phenomenological explanation extending existing theories is presented, that agrees well with our numerical findings. For flows in square ducts, we report four face-centered stable positions for the rotating particles, in accordance with experimental studies. However, the non-rotating particles stay scattered on a concentric one-dimensional annulus, thus revealing qualitatively different behavior with respect to the free ones. Our findings suggest new designs for micro-particle and cell sorting in microfluidics devices.

During social interactions, understanding the intricacies of the context can be vital, particularly for socially anxious individuals. While previous research has found that the presence of a social interaction can be detected from ambient audio, the nuances within social contexts, which influence how anxiety provoking interactions are, remain largely unexplored. As an alternative to traditional, burdensome methods like self-report, this study presents a novel approach that harnesses ambient audio segments to detect social threat contexts. We focus on two key dimensions: number of interaction partners (dyadic vs. group) and degree of evaluative threat (explicitly evaluative vs. not explicitly evaluative). Building on data from a Zoom-based social interaction study (N=52 college students, of whom the majority N=45 are socially anxious), we employ deep learning methods to achieve strong detection performance. Under sample-wide 5-fold Cross Validation (CV), our model distinguished dyadic from group interactions with 90\% accuracy and detected evaluative threat at 83\%. Using a leave-one-group-out CV, accuracies were 82\% and 77\%, respectively. While our data are based on virtual interactions due to pandemic constraints, our method has the potential to extend to diverse real-world settings. This research underscores the potential of passive sensing and AI to differentiate intricate social contexts, and may ultimately advance the ability of context-aware digital interventions to offer personalized mental health support., Comment: 8 pages, 4 figures, 3 tables. Accepted by ACII 2024, Glasgow, UK. To appear in the Proceedings of ACII 2024

I formulate the theory of three dimensional "Malthusian flocks" -- i.e., coherently moving collections of self-propelled entities (such as living creatures) which are being "born" and "dying" during their motion -- whose constituents all have a preference for having their velocity vectors lie parallel to the same two-dimensional plane. I determine the universal scaling exponents characterizing such systems exactly, finding that the dynamical exponent $z=3/2$, the "anisotropy" exponent $\zeta=3/4$, and the "roughness" exponent $\chi=-1/2$. I also give the scaling laws implied by these exponents., Comment: 6 pages, no figures

Tidal interactions between galaxies often give rise to tidal tails, which can harbor concentrations of stars and interstellar gas resembling dwarf galaxies. Some of these tidal dwarf galaxies (TDGs) have the potential to detach from their parent galaxies and become independent entities, but their long-term survival is uncertain. In this study, we conducted a search for detached TDGs associated with a sample of 39 interacting galaxy pairs in the local Universe using infrared, ultraviolet, and optical images. We employed IR colors and UV/optical/IR spectral energy distributions to identify potential interlopers, such as foreground stars or background quasars. Through spectroscopic observations using the Boller and Chivens spectrograph at San Pedro M\'artir Observatory, we confirmed that six candidate TDGs are at the same redshift as their putative parent galaxy pairs. We identified and measured emission lines in the optical spectra and calculated nebular oxygen abundances, which range from log(O/H) = 8.10 $\pm$ 0.01 to 8.51 $\pm$ 0.02. We have serendipitously discovered an additional detached TDG candidate in Arp72 using available spectra from SDSS. Utilizing the photometric data and the CIGALE code for stellar population and dust emission fitting, we derived the stellar masses, stellar population ages, and stellar metallicities for these detached TDGs. Compared to standard mass-metallicity relations for dwarf galaxies, five of the seven candidates have higher than expected metallicities, confirming their tidal origins. One of the seven candidates remains unclear due to large uncertainties in metallicity, and another has stellar and nebular metallicities compatible with those of a preexisting dwarf galaxy. The latter object is relatively compact in the optical relative to its stellar mass, in contrast to the other candidate TDGs [abridged]., Comment: Accepted for publication in Astronomy and Astrophysics. 15 pages, 9 figures, 6 tables

Despite their simplicity, linear models perform well at time series forecasting, even when pitted against deeper and more expensive models. A number of variations to the linear model have been proposed, often including some form of feature normalisation that improves model generalisation. In this paper we analyse the sets of functions expressible using these linear model architectures. In so doing we show that several popular variants of linear models for time series forecasting are equivalent and functionally indistinguishable from standard, unconstrained linear regression. We characterise the model classes for each linear variant. We demonstrate that each model can be reinterpreted as unconstrained linear regression over a suitably augmented feature set, and therefore admit closed-form solutions when using a mean-squared loss function. We provide experimental evidence that the models under inspection learn nearly identical solutions, and finally demonstrate that the simpler closed form solutions are superior forecasters across 72% of test settings.

Robotic manipulation of deformable linear objects (DLOs) is an active area of research, though emerging applications, like automotive wire harness installation, introduce constraints that have not been considered in prior work. Confined workspaces and limited visibility complicate prior assumptions of multi-robot manipulation and direct measurement of DLO configuration (state). This work focuses on single-arm manipulation of stiff DLOs (StDLOs) connected to form a DLO network (DLON), for which the measurements (output) are the endpoint poses of the DLON, which are subject to unknown dynamics during manipulation. To demonstrate feasibility of output-based control without state estimation, direct input-output dynamics are shown to exist by training neural network models on simulated trajectories. Output dynamics are then approximated with polynomials and found to contain well-known rigid body dynamics terms. A composite model consisting of a rigid body model and an online data-driven residual is developed, which predicts output dynamics more accurately than either model alone, and without prior experience with the system. An adaptive model predictive controller is developed with the composite model for DLON manipulation, which completes DLON installation tasks, both in simulation and with a physical automotive wire harness., Comment: Accepted to IEEE International Conference on Robotics and Automation - ICRA 2024. 7 pages. 4 figures

We show that phase separation in ordered polar active fluids belongs to a new universality class. This describes large collections of self-propelled entities (``flocks"), all spontaneously moving in the same direction, in which attractive interactions (which can be caused by, e.g., autochemotaxis) cause phase separation: the system spontaneously separates into a high density band and a low density band, moving parallel to each other, and to the direction of mean flock motion, at different speeds. The upper critical dimension for this transition is $d_c=5$, in contrast to the well-known $d_c=4$ of equilibrium phase separation. We obtain the large-distance, long-time scaling laws of the velocity and density fluctuations, which are characterized by universal critical correlation length and order parameter exponents $\nu_\perp$, $\nu_\parallel$ and $\beta$ respectively. We calculate these to $\mathcal{O} (\epsilon)$ in a $d=5-\epsilon$ expansion., Comment: 21 pages, 9 figures

Motivated by a number of realizations of long-range interacting systems, including ultra-cold atomic and molecular gases, we study a neutral plasma with power-law interactions longer-ranged than Coulombic. We find that beyond a crossover length, such interactions are universally screened down to a standard Coulomb form in all spatial dimensions. This implies, counter-intuitively, that in two dimensions and below, such a "super-Coulombic" gas is asymptotically Coulombically confining at low temperatures. At higher temperatures, the plasma undergoes a deconfining transition that in two dimensions is the same Kosterlitz-Thouless transition that occurs in a conventional Coulomb gas, but at an elevated temperature that we calculate. We also predict that in contrast, above two dimensions, even when naively the bare potential is confining, there is no confined phase of the plasma at any nonzero temperature. In addition, the super-Coulomb to Coulomb crossover is followed at longer length scales by an unconventional "Debye-Huckel" screening, which leads to faster-than-Coulombic, power-law decay of the screened potential, in contrast to the usual exponentially decaying Yukawa potential. Furthermore, we show that power-law potentials, that fall off more rapidly than Coulomb, are screened down to a shorter-ranged power-law, rather than an exponential Debye-Huckel Yukawa form. We expect these prediction to be testable in simulations, and hope they will inspire experimental studies in various platforms., Comment: 17 pages, 2 figures

We develop the hydrodynamic theory of dry, polar ordered, active matter (``flocking") with autochemotaxis; i.e., self-propelled entities moving in the same direction, each emitting a substance which attracts the others (e.g., ants). We find that sufficiently strong autochemotaxis leads to an instability to phase separation into one high and one low density band. This is very analogous to both equilibrium phase separation, and ``motility induced phase separation" (``MIPS") and can occur in flocks due to any microscopic mechanism (e.g., sufficiently strong attractive interactions) that makes the entities cohere., Comment: 6 pages, 5 figures. arXiv admin note: substantial text overlap with arXiv:2401.09461

Multimodal deep learning, especially vision-language models, have gained significant traction in recent years, greatly improving performance on many downstream tasks, including content moderation and violence detection. However, standard multimodal approaches often assume consistent modalities between training and inference, limiting applications in many real-world use cases, as some modalities may not be available during inference. While existing research mitigates this problem through reconstructing the missing modalities, they unavoidably increase unnecessary computational cost, which could be just as critical, especially for large, deployed infrastructures in industry. To this end, we propose a novel guidance network that promotes knowledge sharing during training, taking advantage of the multimodal representations to train better single-modality models to be used for inference. Real-world experiments in violence detection shows that our proposed framework trains single-modality models that significantly outperform traditionally trained counterparts, while avoiding increases in computational cost for inference., Comment: ICME 2024 Camera Ready. Code is available at https://github.com/zhuokaizhao/multimodal-guidance-network

Training neural network classifiers on datasets with label noise poses a risk of overfitting them to the noisy labels. To address this issue, researchers have explored alternative loss functions that aim to be more robust. The `forward-correction' is a popular approach wherein the model outputs are noised before being evaluated against noisy data. When the true noise model is known, applying the forward-correction guarantees consistency of the learning algorithm. While providing some benefit, the correction is insufficient to prevent overfitting to finite noisy datasets. In this work, we propose an approach to tackling overfitting caused by label noise. We observe that the presence of label noise implies a lower bound on the noisy generalised risk. Motivated by this observation, we propose imposing a lower bound on the training loss to mitigate overfitting. Our main contribution is providing theoretical insights that allow us to approximate the lower bound given only an estimate of the average noise rate. We empirically demonstrate that using this bound significantly enhances robustness in various settings, with virtually no additional computational cost.

Individuals high in social anxiety symptoms often exhibit elevated state anxiety in social situations. Research has shown it is possible to detect state anxiety by leveraging digital biomarkers and machine learning techniques. However, most existing work trains models on an entire group of participants, failing to capture individual differences in their psychological and behavioral responses to social contexts. To address this concern, in Study 1, we collected linguistic data from N=35 high socially anxious participants in a variety of social contexts, finding that digital linguistic biomarkers significantly differ between evaluative vs. non-evaluative social contexts and between individuals having different trait psychological symptoms, suggesting the likely importance of personalized approaches to detect state anxiety. In Study 2, we used the same data and results from Study 1 to model a multilayer personalized machine learning pipeline to detect state anxiety that considers contextual and individual differences. This personalized model outperformed the baseline F1-score by 28.0%. Results suggest that state anxiety can be more accurately detected with personalized machine learning approaches, and that linguistic biomarkers hold promise for identifying periods of state anxiety in an unobtrusive way., Comment: IEEE EMBC 2023

We analytically calculate the scaling exponents of a two-dimensional KPZ-like system: coherently moving incompressible polar active fluids. Using three different renormalization group approximation schemes, we obtain values for the ``roughness" exponent $\chi$ and anisotropy exponent $\zeta$ that are extremely near the known exact results. This implies our prediction for the previously completely unknown dynamic exponent $z$ is quantitatively accurate.

Correctly identifying an individual's social context from passively worn sensors holds promise for delivering just-in-time adaptive interventions (JITAIs) to treat social anxiety disorder. In this study, we present results using passively collected data from a within-subject experiment that assessed physiological response across different social contexts (i.e, alone vs. with others), social phases (i.e., pre- and post-interaction vs. during an interaction), social interaction sizes (i.e., dyadic vs. group interactions), and levels of social threat (i.e., implicit vs. explicit social evaluation). Participants in the study ($N=46$) reported moderate to severe social anxiety symptoms as assessed by the Social Interaction Anxiety Scale ($\geq$34 out of 80). Univariate paired difference tests, multivariate random forest models, and follow-up cluster analyses were used to explore physiological response patterns across different social and non-social contexts. Our results suggest that social context is more reliably distinguishable than social phase, group size, or level of social threat, but that there is considerable variability in physiological response patterns even among these distinguishable contexts. Implications for real-world context detection and deployment of JITAIs are discussed.

I show that non-equilibrium two-dimensional interfaces between three dimensional phase separated fluids exhibit a peculiar "sub-logarithmic" roughness. Specifically, an interface of lateral extent $L$ will fluctuate vertically (i.e., normal to the mean surface orientation) a typical RMS distance $w\equiv\sqrt{\langle |h(\br,t)|^2\rangle} \propto [\ln{(L/a)}]^{1/3}$ (where $a$ is a microscopic length, and $ h(\br,t)$ is the height of the interface at two dimensional position $\br$ at time $t$). In contrast, the roughness of equilibrium two-dimensional interfaces between three dimensional fluids, obeys $w \propto [\ln{(L/a)}]^{1/2}$. The exponent $1/3$ for the active case is exact. In addition, the characteristic time scales $\tau(L)$ in the active case scale according to $\tau(L)\propto L^3 [\ln{(L/a)}]^{1/3}$, in contrast to the simple $\tau(L)\propto L^3$ scaling found in equilibrium systems with conserved densities and no fluid flow., Comment: 4 pages, no figures

We show that dislocations in active 2d smectics with underlying rotational symmetry are always unbound in the presence of noise, meaning the active smectic phase does not exist for non-zero noise in $d=2$. The active smectic phase can, like equilibrium smectics in 2d, be stabilized by applying rotational symmetry breaking fields; however, even in the presence of such fields, active smectics are still much less stable against noise than equilibrium ones, when the symmetry breaking field(s) are weak., Comment: 23 pages, 3 figures

We study the moving phase of two-dimensional (2D) incompressible polar active fluids in the presence of both quenched and annealed disorder. We show that long-range polar order persists even in this defect-ridden two-dimensional system. We obtain the large-distance, long-time scaling laws of the velocity fluctuations using three distinct dynamic renormalization group schemes. These are an uncontrolled one-loop calculation in exactly two dimensions, and two $d=(d_c-\epsilon)$-expansions to $O(\epsilon)$, obtained by two different analytic continuations of our 2D model to higher spatial dimensions: a ``hard" continuation which has $d_c={7\over 3}$, and a ``soft" continuation with $d_c={5\over 2}$. Surprisingly, the quenched and annealed parts of the velocity correlation function have the same anisotropy exponent and the relaxational and propagating parts of the dispersion relation have the same dynamic exponent in the nonlinear theory even though they are distinct in the linearized theory. This is due to anomalous hydrodynamics. Furthermore, all three renormalization schemes yield very similar values for the universal exponents, and, therefore, we expect the numerical values we predict for them to be highly accurate., Comment: 39 pages, 11 figures