We study the natural function space for infinitely wide two-layer neural networks with ReLU activation (Barron space) and establish different representation formulae. In two cases, we describe the space explicitly up ...
详细信息
Large CNNs have delivered impressive performance in various computer vision applications. But the storage and computation requirements make it problematic for deploying these models on mobile devices. Recently, tensor...
详细信息
Multiple myeloma is a hematological malignancy characterized by proliferation of malignant plasma cells and derangement of bone homeostasis. Myeloma bone disease results in significant morbidity as a result of bone pa...
详细信息
ISBN:
(纸本)9783950353709
Multiple myeloma is a hematological malignancy characterized by proliferation of malignant plasma cells and derangement of bone homeostasis. Myeloma bone disease results in significant morbidity as a result of bone pain, hypercalcemia, diffuse osteopenia, and pathologic fractures. We present a spatially explicit mathematical model of multiple myeloma and bone remodeling, synthesizing the existing model of local "microenvironment" interactions in Ayati et al. 2010 [1] with a level set approach for representing the sharp interface been bone and marrow introduced in [6]. computational results show the feasibility of using a level set to capture the spatial structure in the context of a geometrically straightforward interface, but one that nonetheless captures the essence of the rich geometries seen in bone marrow biopsy slides. In particular, we are able to model the formation of an osteolytic lesion in the case of multiple myeloma dysregulated bone remodeling, but not, using the same remodeling parameter set, in the case of normal bone remodeling.
In our previous work [1], we studied an interconnected bursting neuron model for insect locomotion, and its corresponding phase oscillator model, which at high speed can generate stable tripod gaits with three legs of...
详细信息
We report on an extensive study of the viscosity of liquid water at near-ambient conditions,performed within the Green-Kubo theory of linear response and equilibrium ab initio molecular dynamics(AIMD),based on density...
详细信息
We report on an extensive study of the viscosity of liquid water at near-ambient conditions,performed within the Green-Kubo theory of linear response and equilibrium ab initio molecular dynamics(AIMD),based on density-functional theory(DFT).In order to cope with the long simulation times necessary to achieve an acceptable statistical accuracy,our ab initio approach is enhanced with deep-neural-network potentials(NNP).This approach is first validated against AIMD results,obtained by using the Perdew–Burke–Ernzerhof(PBE)exchange-correlation functional and paying careful attention to crucial,yet often overlooked,aspects of the statistical data ***,we train a second NNP to a dataset generated from the Strongly Constrained and Appropriately Normed(SCAN)*** the error resulting from the imperfect prediction of the melting line is offset by referring the simulated temperature to the theoretical melting one,our SCAN predictions of the shear viscosity of water are in very good agreement with experiments.
This paper presents a novel algorithm for the 3D tomographic inversion problem that arises in single-particle electron cryomicroscopy (Cryo-EM). It is based on two key components: 1) a variational formulation that pro...
详细信息
This paper presents a novel algorithm for the 3D tomographic inversion problem that arises in single-particle electron cryomicroscopy (Cryo-EM). It is based on two key components: 1) a variational formulation that promotes sparsity in the wavelet domain and 2) the Toeplitz structure of the combined projection/back-projection operator. The first idea has proven to be very effective for the recovery of piecewise-smooth signals, which is confirmed by our numerical experiments. The second idea allows for a computationally efficient implementation of the reconstruction procedure, using only one circulant convolution per iteration.
Machine learning models are changing the paradigm of molecular modeling, which is a fundamental tool for material science, chemistry, and computational biology. Of particular interest is the inter-atomic potential ene...
Machine learning models are changing the paradigm of molecular modeling, which is a fundamental tool for material science, chemistry, and computational biology. Of particular interest is the inter-atomic potential energy surface (PES). Here we develop Deep Potential - Smooth Edition (DeepPot-SE), an end-to-end machine learning-based PES model, which is able to efficiently represent the PES of a wide variety of systems with the accuracy of ab initio quantum mechanics models. By construction, DeepPot-SE is extensive and continuously differentiable, scales linearly with system size, and preserves all the natural symmetries of the system. Further, we show that DeepPot-SE describes finite and extended systems including organic molecules, metals, semiconductors, and insulators with high fidelity.
Osteocytes regulate the response of osteoclasts and osteoblasts to mechanical loading through signaling molecules, the levels of which are controlled by post-translational modification or degradation and by differenti...
Osteocytes regulate the response of osteoclasts and osteoblasts to mechanical loading through signaling molecules, the levels of which are controlled by post-translational modification or degradation and by differential gene transcription and translation. The magnitude and mode of bone tissue deformation that elicits a transcriptional response in individual osteocytes in situ has been difficult to quantify. We measured SOST, Wnt11, TNF, and FRZB gene expression in osteocytes within loaded and unloaded control porcine trabecular bone explants using RNAScope® and compared the local tissue level strain and strain gradient-which we used as an indicator of potential poroelastic fluid flow-in the tissue surrounding osteocytes with high vs. low gene expression. The measured expression of all four genes differed between loaded and unloaded explants, on average, with the mean SOST expression level decreasing by 45%. In the loaded explants, gene expression was altered from baseline in about 30% of the osteocytes, and they were surrounded by tissue with higher strain and strain gradient than the 20 to 25% of osteocytes that remained near baseline expression. Both deviatoric strain and hydrostatic strain gradient were sensitive and specific predictors of the mechanobiological response for individual genes as well as combinations. SOST expression was highly related to elevated strain gradient, providing evidence that osteocytes respond to fluid flow in the lacunar-canalicular system.
We consider semidefinite programs (SDPs) with equality constraints. The variable to be optimized is a positive semidefinite matrix X of size n. Following the Burer–Monteiro approach, we optimize a factor Y of size n&...
详细信息
An active learning procedure called deep potential generator (DP-GEN) is proposed for the construction of accurate and transferable machine learning-based models of the potential energy surface (PES) for the molecular...
详细信息
An active learning procedure called deep potential generator (DP-GEN) is proposed for the construction of accurate and transferable machine learning-based models of the potential energy surface (PES) for the molecular modeling of materials. This procedure consists of three main components: exploration, generation of accurate reference data, and training. Application to the sample systems of Al, Mg, and Al-Mg alloys demonstrates that DP-GEN can produce uniformly accurate PES models with a minimal number of reference data.
暂无评论