Search

The evolution in latitude of sunspots is a key feature of the cyclic solar dynamo. Here, we present the results of a spectroscopic and spectropolarimetric monitoring campaign on the young (∼20 Myr old) early G star HIP 89829, in order to investigate potential evolution in the distribution of the star's spots and magnetic features. Our analysis of this G5V star spans eight epochs, from June 2010 to August 2015. The techniques of Doppler imaging and Zeeman–Doppler imaging were used to create brightness maps for each epoch and magnetic maps for two epochs. The brightness images show the star to have stable spot features with two main spot latitudes – a polar spot, often seen on young rapidly rotating stars such as this, and another highly unusual group of large spot features around the 20° and 30° latitudes. These lower spot latitudes appear to be rather stable over the 5 yr of observations. We included a solar-type differential rotation law into the imaging process and measured near-solid-body rotation for epochs where sufficient data exist for this analysis. The magnetic features show a dominant poloidal and a weaker toroidal magnetic field for both Stokes V epochs, which is unusual for a star with a rapid rotation period of 0.57 d. We conclude that HIP 89829 is an active young solar-type star with long-lived spots and near-solid-body rotation. [ABSTRACT FROM AUTHOR]

The Hopfield model is a paradigmatic model of neural networks that has been analyzed for many decades in the statistical physics, neuroscience, and machine learning communities. Inspired by the manifold hypothesis in machine learning, we propose and investigate a generalization of the standard setting that we name random-features Hopfield model. Here, P binary patterns of length N are generated by applying to Gaussian vectors sampled in a latent space of dimension D a random projection followed by a nonlinearity. Using the replica method from statistical physics, we derive the phase diagram of the model in the limit P,N,D→∞ with fixed ratios α=P/N and α_{D}=D/N. Besides the usual retrieval phase, where the patterns can be dynamically recovered from some initial corruption, we uncover a new phase where the features characterizing the projection can be recovered instead. We call this phenomena the learning phase transition, as the features are not explicitly given to the model but rather are inferred from the patterns in an unsupervised fashion.

Empirical studies on the landscape of neural networks have shown that low-energy configurations are often found in complex connected structures, where zero-energy paths between pairs of distant solutions can be constructed. Here, we consider the spherical negative perceptron, a prototypical nonconvex neural network model framed as a continuous constraint satisfaction problem. We introduce a general analytical method for computing energy barriers in the simplex with vertex configurations sampled from the equilibrium. We find that in the overparametrized regime the solution manifold displays simple connectivity properties. There exists a large geodesically convex component that is attractive for a wide range of optimization dynamics. Inside this region we identify a subset of atypical high-margin solutions that are geodesically connected with most other solutions, giving rise to a star-shaped geometry. We analytically characterize the organization of the connected space of solutions and show numerical evidence of a transition, at larger constraint densities, where the aforementioned simple geodesic connectivity breaks down.

Prostate cancer (PC) is the most frequently diagnosed cancer among adult men, and its incidence is increasing worldwide [...].

We study the binary and continuous negative-margin perceptrons as simple nonconvex neural network models learning random rules and associations. We analyze the geometry of the landscape of solutions in both models and find important similarities and differences. Both models exhibit subdominant minimizers which are extremely flat and wide. These minimizers coexist with a background of dominant solutions which are composed by an exponential number of algorithmically inaccessible small clusters for the binary case (the frozen 1-RSB phase) or a hierarchical structure of clusters of different sizes for the spherical case (the full RSB phase). In both cases, when a certain threshold in constraint density is crossed, the local entropy of the wide flat minima becomes nonmonotonic, indicating a breakup of the space of robust solutions into disconnected components. This has a strong impact on the behavior of algorithms in binary models, which cannot access the remaining isolated clusters. For the spherical case the behavior is different, since even beyond the disappearance of the wide flat minima the remaining solutions are shown to always be surrounded by a large number of other solutions at any distance, up to capacity. Indeed, we exhibit numerical evidence that algorithms seem to find solutions up to the SAT/UNSAT transition, that we compute here using an 1RSB approximation. For both models, the generalization performance as a learning device is shown to be greatly improved by the existence of wide flat minimizers even when trained in the highly underconstrained regime of very negative margins.

Millions of liters of diluted bitumen (dilbit), a crude oil product from Canada's oil sands region, is transported through critical Pacific salmon habitat each day. While the toxicity of the water-soluble fraction of dilbit (WSFd) to early life-stages of salmon is known, quantitative data on life-stage differences in sensitivity to WSFd is missing. To fill this knowledge gap, we exposed two juvenile life-stages of coho salmon (O. kisutch) in parallel to very low (parts per billion), environmentally-relevant concentrations of WSFd for acute (48 h) and sub-chronic (4 wk) durations. The relative sensitivities of the two life-stages (fry and parr) were assessed by comparing the timing and magnitude of biological responses using common organismal and molecular endpoints of crude oil exposure. A significant reduction in body condition occurred in both fry and parr after 4 wk exposure to WSFd. Both life-stages also experienced a concentration-dependent decrease in time-to-loss-of-equilibrium during a hypoxia challenge test at both 48 h and 4 wk of exposure. Although organismal responses were similar, molecular responses were distinct between life-stages. In general, unexposed fry had higher baseline values of hepatic phase I biotransformation indicators than unexposed parr, but induction of EROD activity and cyp1a mRNA expression in response to WSFd exposure was greater in parr than in fry. Neither gst nor hsp70 mRNA expression, markers of phase II biotransformation and cell stress, respectively, were reliably altered by WSFd exposure in either life-stage. Taken together, results of this study do not support differential sensitivities of coho fry and parr to WSFd. All the same, the potential for ontogenic differences in the expression and induction of phase I biotransformation need to be considered because age does matter for these endpoints if they are used as bioindicators of exposure in post-spill impact assessments., Competing Interests: Declaration of Competing Interest The authors declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper., (Copyright © 2022. Published by Elsevier B.V.)

PI-RADS 3 prostate lesions clinical management is still debated, with high variability among different centers. Identifying clinically significant tumors among PI-RADS 3 is crucial. Radiomics applied to multiparametric MR (mpMR) seems promising. Nevertheless, reproducibility assessment by external validation is required. We retrospectively included all patients with at least one PI-RADS 3 lesion (PI-RADS v2.1) detected on a 3T prostate MRI scan at our Institution (June 2016-March 2021). An MRI-targeted biopsy was used as ground truth. We assessed reproducible mpMRI radiomic features found in the literature. Then, we proposed a new model combining PSA density and two radiomic features (texture regularity (T2) and size zone heterogeneity (ADC)). All models were trained/assessed through 100-repetitions 5-fold cross-validation. Eighty patients were included (26 with GS ≥ 7). In total, 9/20 T2 features (Hector's model) and 1 T2 feature (Jin's model) significantly correlated to biopsy on our dataset. PSA density alone predicted clinically significant tumors (sensitivity: 66%; specificity: 71%). Our model obtained a sensitivity of 80% and a specificity of 76%. Standard-compliant works with detailed methodologies achieve comparable radiomic feature sets. Therefore, efforts to facilitate reproducibility are needed, while complex models and imaging protocols seem not, since our model combining PSA density and two radiomic features from routinely performed sequences appeared to differentiate clinically significant cancers.

Current deep neural networks are highly overparameterized (up to billions of connection weights) and nonlinear. Yet they can fit data almost perfectly through variants of gradient descent algorithms and achieve unexpected levels of prediction accuracy without overfitting. These are formidable results that defy predictions of statistical learning and pose conceptual challenges for nonconvex optimization. In this paper, we use methods from statistical physics of disordered systems to analytically study the computational fallout of overparameterization in nonconvex binary neural network models, trained on data generated from a structurally simpler but "hidden" network. As the number of connection weights increases, we follow the changes of the geometrical structure of different minima of the error loss function and relate them to learning and generalization performance. A first transition happens at the so-called interpolation point, when solutions begin to exist (perfect fitting becomes possible). This transition reflects the properties of typical solutions, which however are in sharp minima and hard to sample. After a gap, a second transition occurs, with the discontinuous appearance of a different kind of "atypical" structures: wide regions of the weight space that are particularly solution dense and have good generalization properties. The two kinds of solutions coexist, with the typical ones being exponentially more numerous, but empirically we find that efficient algorithms sample the atypical, rare ones. This suggests that the atypical phase transition is the relevant one for learning. The results of numerical tests with realistic networks on observables suggested by the theory are consistent with this scenario.

The success of deep learning has revealed the application potential of neural networks across the sciences and opened up fundamental theoretical problems. In particular, the fact that learning algorithms based on simple variants of gradient methods are able to find near-optimal minima of highly nonconvex loss functions is an unexpected feature of neural networks. Moreover, such algorithms are able to fit the data even in the presence of noise, and yet they have excellent predictive capabilities. Several empirical results have shown a reproducible correlation between the so-called flatness of the minima achieved by the algorithms and the generalization performance. At the same time, statistical physics results have shown that in nonconvex networks a multitude of narrow minima may coexist with a much smaller number of wide flat minima, which generalize well. Here, we show that wide flat minima arise as complex extensive structures, from the coalescence of minima around "high-margin" (i.e., locally robust) configurations. Despite being exponentially rare compared to zero-margin ones, high-margin minima tend to concentrate in particular regions. These minima are in turn surrounded by other solutions of smaller and smaller margin, leading to dense regions of solutions over long distances. Our analysis also provides an alternative analytical method for estimating when flat minima appear and when algorithms begin to find solutions, as the number of model parameters varies.