Unsupervised learning with generative adversarial networks (GANs) has proven hugely successful. Regular GANs hypothesize the discriminator as a classifier with the sigmoid cross entropy loss function. However, we found that this loss function may lead to the vanishing gradients problem during the learning process. To overcome such a problem, we propose in this paper the Least Squares Generative Adversarial Networks (LS-GANs) which adopt the least squares loss function for the discriminator. We show that minimizing the objective function of LSGAN yields minimizing the Pearson χ 2 divergence. There are two benefits of LSGANs over regular GANs. First, LSGANs are able to generate higher quality images than regular GANs. Second, LSGANs perform more stable during the learning process. We evaluate LSGANs on five scene datasets and the experimental results show that the images generated by LSGANs are of better quality than the ones generated by regular GANs. We also conduct two comparison experiments between LSGANs and regular GANs to illustrate the stability of LSGANs.

State space models (SSMs) have demonstrated state-of-the-art sequence modeling performance in some modalities, but underperform attention in language modeling. Moreover, despite scaling nearly linearly in sequence length instead of quadratically, SSMs are still slower than Transformers due to poor hardware utilization. In this paper, we make progress on understanding the expressivity gap between SSMs and attention in language modeling, and on reducing the hardware barrier between SSMs and attention. First, we use synthetic language modeling tasks to understand the gap between SSMs and attention. We find that existing SSMs struggle with two capabilities: recalling earlier tokens in the sequence and comparing tokens across the sequence. To understand the impact on language modeling, we propose a new SSM layer, H3, that is explicitly designed for these abilities. H3 matches attention on the synthetic languages and comes within 0.4 PPL of Transformers on OpenWebText. Furthermore, a hybrid 125M-parameter H3-attention model that retains two attention layers surprisingly outperforms Transformers on OpenWebText by 1.0 PPL. Next, to improve the efficiency of training SSMs on modern hardware, we propose FlashConv. FlashConv uses a fused block FFT algorithm to improve efficiency on sequences up to 8K, and introduces a novel state passing algorithm that exploits the recurrent properties of SSMs to scale to longer sequences. FlashConv yields 2$\times$ speedup on the long-range arena benchmark and allows hybrid language models to generate text 1.6$\times$ faster than Transformers. Using FlashConv, we scale hybrid H3-attention language models up to 1.3B parameters on the Pile and find promising initial results, achieving lower perplexity than Transformers and outperforming Transformers in zero- and few-shot learning on a majority of tasks in the SuperGLUE benchmark.

A "heart attack" or myocardial infarction (MI), occurs when an artery supplying blood to the heart is abruptly occluded. The "gold standard" method for imaging MI is Cardiovascular Magnetic Resonance Imaging (MRI), with intravenously administered gadolinium-based contrast (late gadolinium enhancement). However, no "gold standard" fully automated method for the quantification of MI exists. In this work, we propose an end-to-end fully automatic system (MyI-Net) for the detection and quantification of MI in MRI images. This has the potential to reduce the uncertainty due to the technical variability across labs and inherent problems of the data and labels. Our system consists of four processing stages designed to maintain the flow of information across scales. First, features from raw MRI images are generated using feature extractors built on ResNet and MoblieNet architectures. This is followed by the Atrous Spatial Pyramid Pooling (ASPP) to produce spatial information at different scales to preserve more image context. High-level features from ASPP and initial low-level features are concatenated at the third stage and then passed to the fourth stage where spatial information is recovered via up-sampling to produce final image segmentation output into: i) background, ii) heart muscle, iii) blood and iv) scar areas. New models were compared with state-of-art models and manual quantification. Our models showed favorable performance in global segmentation and scar tissue detection relative to state-of-the-art work, including a four-fold better performance in matching scar pixels to contours produced by clinicians.

In the Earth's magnetosphere, there are fewer than a dozen dedicated probes beyond low-Earth orbit making in-situ observations at any given time. As a result, we poorly understand its global structure and evolution, the mechanisms of its main activity processes, magnetic storms, and substorms. New Artificial Intelligence (AI) methods, including machine learning, data mining, and data assimilation, as well as new AI-enabled missions will need to be developed to meet this Sparse Data challenge.

Various graph neural networks (GNNs) have been proposed to solve node classification tasks in machine learning for graph data. GNNs use the structural information of graph data by aggregating the features of neighboring nodes. However, they fail to directly characterize and leverage the structural information. In this paper, we propose multi-duplicated characterization of graph structures using information gain ratio (IGR) for GNNs (MSI-GNN), which enhances the performance of node classification by using an i-hop adjacency matrix as the structural information of the graph data. In MSI-GNN, the i-hop adjacency matrix is adaptively adjusted by two methods: (i) structural features in the matrix are selected based on the IGR, and (ii) the selected features in (i) for each node are duplicated and combined flexibly. In an experiment, we show that our MSI-GNN outperforms GCN, H2GCN, and GCNII in terms of average accuracies in benchmark graph datasets.

Weakly-supervised learning (WSL) has been proposed to alleviate the conflict between data annotation cost and model performance through employing sparsely-grained (i.e., point-, box-, scribble-wise) supervision and has shown promising performance, particularly in the image segmentation field. However, it is still a very challenging problem due to the limited supervision, especially when only a small number of labeled samples are available. Additionally, almost all existing WSL segmentation methods are designed for star-convex structures which are very different from curvilinear structures such as vessels and nerves. In this paper, we propose a novel sparsely annotated segmentation framework for curvilinear structures, named YoloCurvSeg, based on image synthesis. A background generator delivers image backgrounds that closely match real distributions through inpainting dilated skeletons. The extracted backgrounds are then combined with randomly emulated curves generated by a Space Colonization Algorithm-based foreground generator and through a multilayer patch-wise contrastive learning synthesizer. In this way, a synthetic dataset with both images and curve segmentation labels is obtained, at the cost of only one or a few noisy skeleton annotations. Finally, a segmenter is trained with the generated dataset and possibly an unlabeled dataset. The proposed YoloCurvSeg is evaluated on four publicly available datasets (OCTA500, CORN, DRIVE and CHASEDB1) and the results show that YoloCurvSeg outperforms state-of-the-art WSL segmentation methods by large margins. With only one noisy skeleton annotation (respectively 0.14%, 0.02%, 1.4%, and 0.65% of the full annotation), YoloCurvSeg achieves more than 97% of the fully-supervised performance on each dataset. Code and datasets will be released at https://github.com/llmir/YoloCurvSeg.

开发有效的自动分类器将真实来源与工件分开，对于宽场光学调查的瞬时随访至关重要。在图像差异过程之后，从减法伪像的瞬态检测鉴定是此类分类器的关键步骤，称为真实 - 博格斯分类问题。我们将自我监督的机器学习模型，深入的自组织地图（DESOM）应用于这个“真实的模拟”分类问题。 DESOM结合了自动编码器和一个自组织图以执行聚类，以根据其维度降低的表示形式来区分真实和虚假的检测。我们使用32x32归一化检测缩略图作为底部的输入。我们展示了不同的模型训练方法，并发现我们的最佳DESOM分类器显示出6.6％的检测率，假阳性率为1.5％。 Desom提供了一种更细微的方法来微调决策边界，以确定与其他类型的分类器（例如在神经网络或决策树上构建的）结合使用时可能进行的实际检测。我们还讨论了DESOM及其局限性的其他潜在用法。

现有的数据驱动和反馈流量控制策略不考虑实时数据测量的异质性。此外，对于缺乏数据效率，传统的加固学习方法（RL）方法通常会缓慢收敛。此外，常规的最佳外围控制方案需要对系统动力学的精确了解，因此对内源性不确定性会很脆弱。为了应对这些挑战，这项工作提出了一种基于不可或缺的增强学习（IRL）的方法来学习宏观交通动态，以进行自适应最佳周边控制。这项工作为运输文献做出了以下主要贡献：（a）开发连续的时间控制，并具有离散增益更新以适应离散时间传感器数据。 （b）为了降低采样复杂性并更有效地使用可用数据，将体验重播（ER）技术引入IRL算法。 （c）所提出的方法以“无模型”方式放松模型校准的要求，该方式可以稳健地进行建模不确定性，并通过数据驱动的RL算法增强实时性能。 （d）通过Lyapunov理论证明了基于IRL的算法和受控交通动力学的稳定性的收敛性。最佳控制定律被参数化，然后通过神经网络（NN）近似，从而缓解计算复杂性。在不需要模型线性化的同时，考虑了状态和输入约束。提出了数值示例和仿真实验，以验证所提出方法的有效性和效率。

ICECUBE是一种用于检测1 GEV和1 PEV之间大气和天体中微子的光学传感器的立方公斤阵列，该阵列已部署1.45 km至2.45 km的南极的冰盖表面以下1.45 km至2.45 km。来自ICE探测器的事件的分类和重建在ICeCube数据分析中起着核心作用。重建和分类事件是一个挑战，这是由于探测器的几何形状，不均匀的散射和冰中光的吸收，并且低于100 GEV的光，每个事件产生的信号光子数量相对较少。为了应对这一挑战，可以将ICECUBE事件表示为点云图形，并将图形神经网络（GNN）作为分类和重建方法。 GNN能够将中微子事件与宇宙射线背景区分开，对不同的中微子事件类型进行分类，并重建沉积的能量，方向和相互作用顶点。基于仿真，我们提供了1-100 GEV能量范围的比较与当前ICECUBE分析中使用的当前最新最大似然技术，包括已知系统不确定性的影响。对于中微子事件分类，与当前的IceCube方法相比，GNN以固定的假阳性速率（FPR）提高了信号效率的18％。另外，GNN在固定信号效率下将FPR的降低超过8（低于半百分比）。对于能源，方向和相互作用顶点的重建，与当前最大似然技术相比，分辨率平均提高了13％-20％。当在GPU上运行时，GNN能够以几乎是2.7 kHz的中位数ICECUBE触发速率的速率处理ICECUBE事件，这打开了在在线搜索瞬态事件中使用低能量中微子的可能性。

电子显微镜数据集的自动分析提出了多个挑战，例如训练数据集规模的限制，样品质量和实验条件变化引起的数据分布的变化等。受过训练的模型继续提供可接受的细分/新数据上的分类性能，并量化与其预测相关的不确定性。在机器学习的广泛应用中，已经采用了各种方法来量化不确定性，例如贝叶斯建模，蒙特卡洛辍学，合奏等。目的是解决电子显微镜数据域特有的挑战，两种不同类型的类型这项工作实施了预训练的神经网络的合奏。合奏在两相混合物中对冰晶进行语义分割，从而跟踪其相变成水。第一个合奏（EA）由具有不同基础体系结构的U-NET样式网络组成，而第二系列合奏（ER-I）由随机初始化的U-NET样式网络组成，每个基础学习者都具有相同的基础架构'一世'。基础学习者的编码者已在Imagenet数据集上进行了预训练。对EA和ER的性能进行了三个不同的指标评估：准确性，校准和不确定性。可以看出，与ER相比，EA具有更高的分类精度，并且可以更好地校准。尽管这两种类型的集合的不确定性量化是可比的，但ER所表现出的不确定性得分依赖于其基本成员（'i'）的特定架构，并且不一致地比EA更好。因此，与像ER这样的合奏设计相比，像EA这样的合奏设计对电子显微镜数据集的分析所带来的挑战似乎可以更好地解决。