LRGCPND: Predicting Associations between ncRNA and Drug Resistance via Linear Residual Graph Convolution

Accurate inference of the relationship between non-coding RNAs (ncRNAs) and drug resistance is essential for understanding the complicated mechanisms of drug actions and clinical treatment. Traditional biological experiments are time-consuming, laborious, and minor in scale. Although several databases provide relevant resources, computational method for predicting this type of association has not yet been developed. In this paper, we leverage the verified association data of ncRNA and drug resistance to construct a bipartite graph and then develop a linear residual graph convolution approach for predicting associations between non-coding RNA and drug resistance (LRGCPND) without introducing or defining additional data. LRGCPND first aggregates the potential features of neighboring nodes per graph convolutional layer. Next, we transform the information between layers through a linear function. Eventually, LRGCPND unites the embedding representations of each layer to complete the prediction. Results of comparison experiments demonstrate that LRGCPND has more reliable performance than seven other state-of-the-art approaches with an average AUC value of 0.8987. Case studies illustrate that LRGCPND is an effective tool for inferring the associations between ncRNA and drug resistance.

Fast Flow Reconstruction via Robust Invertible n×n Convolution

Flow-based generative models have recently become one of the most efficient approaches to model data generation. Indeed, they are constructed with a sequence of invertible and tractable transformations. Glow first introduced a simple type of generative flow using an invertible 1×1 convolution. However, the 1×1 convolution suffers from limited flexibility compared to the standard convolutions. In this paper, we propose a novel invertible n×n convolution approach that overcomes the limitations of the invertible 1×1 convolution. In addition, our proposed network is not only tractable and invertible but also uses fewer parameters than standard convolutions. The experiments on CIFAR-10, ImageNet and Celeb-HQ datasets, have shown that our invertible n×n convolution helps to improve the performance of generative models significantly.

On the escape of a resonantly excited couple of particles from a potential well

The escape dynamics of a damped system of two coupled particles in a truncated potential well under biharmonic excitation are investigated. It is assumed that excitation frequencies are tuned to the modal natural frequency of the relative motion and to the modal frequency of the centre of mass on the bottom of the potential well. Although the escape is essentially a non-stationary process, the critical force strongly depends on the stationary amplitude of the relative vibrations within the pair of masses. The characteristic escape curve for the critical force moves up on the frequency-escape threshold plane with increasing relative vibrations, which can be interpreted as a stabilizing effect due to the high-frequency excitation. To obtain the results, new modelling techniques are suggested, including the reduction in the effect of the high-frequency excitation using a probability density function-based convolution approach and an energy-based approach for the description of the evolution of the slow variables. To validate the method, the coupled pair of particles is investigated with various model potentials.

Implementation of geostatistical models for large spatiotemporal datasets using multi-resolution approximations

<p>The multi-resolution approximation approach (MRA) [1] provides an efficient representation of Gaussian processes that scales beyond millions of observations. MRA leaves flexibility in the selection of covariance functions and allows to trade off computation time against prediction performance, depending on the selection of parameters. Recent work [2] has shown how MRA can be used for global spatiotemporal processes by integrating nonstationary covariance functions, where parameters vary over space and/or time following a kernel convolution approach. As such, MRA turns out to be a promising approach for geostatistical modelling of global spatiotemporal datasets, such as those coming from Earth observation satellites.</p><p>In this work, we show how MRA can be used for spatiotemporal analysis from a practical perspective. In the first part, we will discuss the influence of parameters (spatiotemporal shape of partitioning regions, the number of basis functions, and the number of partitioning levels) by analyzing a real world dataset. In the second part, we will present and discuss our implementation as an R package stmra[3]. We will demonstrate how traditional models as from the gstat package can be implemented efficiently with MRA, and how non-stationary models can be defined by users in a relatively simple way. </p><p>[1] Katzfuss, M. (2017). A multi-resolution approximation for massive spatial datasets. Journal of the American Statistical Association, 112(517), 201-214</p><p>[2] Appel, M., & Pebesma, E. (2020). Spatiotemporal multi-resolution approximations for analyzing global environmental data. Spatial Statistics, 38, 100465.</p><p>[3] https://github.com/appelmar/stmra</p>

A Spherical Convolution Approach for Learning Long Term Viewport Prediction in 360 Immersive Video

Viewport prediction for 360 video forecasts a viewer’s viewport when he/she watches a 360 video with a head-mounted display, which benefits many VR/AR applications such as 360 video streaming and mobile cloud VR. Existing studies based on planar convolutional neural network (CNN) suffer from the image distortion and split caused by the sphere-to-plane projection. In this paper, we start by proposing a spherical convolution based feature extraction network to distill spatial-temporal 360 information. We provide a solution for training such a network without a dedicated 360 image or video classification dataset. We differ with previous methods, which base their predictions on image pixel-level information, and propose a semantic content and preference based viewport prediction scheme. In this paper, we adopt a recurrent neural network (RNN) network to extract a user's personal preference of 360 video content from minutes of embedded viewing histories. We utilize this semantic preference as spatial attention to help network find the "interested'' regions on a future video. We further design a tailored mixture density network (MDN) based viewport prediction scheme, including viewport modeling, tailored loss function, etc, to improve efficiency and accuracy. Our extensive experiments demonstrate the rationality and performance of our method, which outperforms state-of-the-art methods, especially in long-term prediction.

PCDRN: Progressive Cascade Deep Residual Network for Pansharpening

Pansharpening is the process of fusing a low-resolution multispectral (LRMS) image with a high-resolution panchromatic (PAN) image. In the process of pansharpening, the LRMS image is often directly upsampled by a scale of 4, which may result in the loss of high-frequency details in the fused high-resolution multispectral (HRMS) image. To solve this problem, we put forward a novel progressive cascade deep residual network (PCDRN) with two residual subnetworks for pansharpening. The network adjusts the size of an MS image to the size of a PAN image twice and gradually fuses the LRMS image with the PAN image in a coarse-to-fine manner. To prevent an overly-smooth phenomenon and achieve high-quality fusion results, a multitask loss function is defined to train our network. Furthermore, to eliminate checkerboard artifacts in the fusion results, we employ a resize-convolution approach instead of transposed convolution for upsampling LRMS images. Experimental results on the Pléiades and WorldView-3 datasets prove that PCDRN exhibits superior performance compared to other popular pansharpening methods in terms of quantitative and visual assessments.