--Language models are at the heart of numerous works, notably in the text mining and information retrieval communities. These statistical models aim at extracting word distributions, from simple unigram models to recurrent approaches with latent variables that capture subtle dependencies in texts. However, those models are learned from word sequences only, and authors' identities, as well as publication dates, are seldom considered. We propose a neural model, based on recurrent language modeling, which aims at capturing language diffusion tendencies in author communities through time. By conditioning language models with author and temporal vector states, we are able to leverage the latent dependencies between the text contexts. This allows us to beat several temporal and non-temporal language baselines on two real-world corpora, and to learn meaningful author representations that vary through time. Index T erms --representation learning, dynamic language model, diachronic text analysis I. I NTRODUCTION Language modeling has been at the heart of a huge amount of works for decades. While the natural language processing field focuses on fine-grained text analysis, statistical models for information retrieval and text mining are essentially based on word (or N-gram) counts, considering more or less complex dependencies in texts. Early works in this area focused on the unigram multinomial model [1], and recent works are shifting toward neural approaches, with distributed representations of words [2], [3]. Research on these deep language models is very active [4]-[8], with applications in various text-related tasks such as speech recognition [9], image captioning [10], or text generation [11]. The goal of the language modeling task is to determine word distributions, depending on their context. Classically, these contexts are limited to previous or surrounding words in text documents. However, textual documents often come with additional contextual information, namely their authors and publication dates. Leveraging this additional contextual information is thus a key challenge in order to build more efficient language models.

Graph Convolutional Neural Networks (GCNNs) extend classical CNNs to graph data domain, such as brain networks, social networks and 3D point clouds. It is critical to identify an appropriate graph for the subsequent graph convolution. Existing methods manually construct or learn one fixed graph for all the layers of a GCNN. In order to adapt to the underlying structure of node features in different layers, we propose dynamic learning of graphs and node features jointly in GCNNs. In particular, we cast the graph optimization problem as distance metric learning to capture pairwise similarities of features in each layer. We deploy the Mahalanobis distance metric and further decompose the metric matrix into a low-dimensional matrix, which converts graph learning to the optimization of a low-dimensional matrix for efficient implementation. Extensive experiments on point clouds and citation network datasets demonstrate the superiority of the proposed method in terms of both accuracies and robustness.

Abstract: Recent automated crop mapping via supervised le arning - based methods have demonstrated unprecedented improvement over classical techniques. However, m ost crop mapping studies are limited to same - year crop mapping in which the present year's labeled data is used to predict the same year's crop map. Cross - y ear crop mapping is more useful as it allows the prediction of the following years' crop maps using previously labeled data. We propose Vector Dynamic Time Warping ( VD TW), a novel multi - year classification approach based on warping of angular distances between phenological vectors. The results prove that the proposed VDTW method is robust to temporal and spectral v ariations compensating for different farming practices, climate and atmospheric effects, and measurement errors between years. We also describe a method for determining the most discriminative time window that allows high classification accuracies with lim ited data. We carried out test s of our approach with Lan dsat 8 time - series imagery from years 2013 to 2016 for classification of corn and cotton in the Harran Plain, and corn, cotton, and soybean in the Bismil Plain of Southeastern Turkey. In addition, we tested VDTW corn and soybean in Kansas, the US for 2017 and 2018 with the Harmonized Landsat Sentinel data . The VDTW method achieved 99.85% and 99.74% overall accuracies for the same and cross years, respectively with fewer training samples compared to oth er state - of - the - art approaches, i.e. spectral angle mapp er ( SAM), dynamic time warping ( DTW), time - weighted DTW ( TWDTW), random forest (RF), support vector machine ( SVM) and deep long short - term memory ( LSTM) methods. The proposed method could be expanded for other crop types and/or geographical areas. Keywords: Time series; phenology; multi - year classification; dynamic programming; Landsat; crop mapping; land use; corn; cotton; soybean 1. Introduction T he world population is expected to exceed nine billion in 2050 [1] . Providing adequate nutrition for the increasing human population is a significant concern. Advanced agri cultural technologies, such as precision agriculture and precision irrigation are rapidly emerging to optimize water, fertilizers, and pesticides; thereby enabling higher crop yield. Accurate crop maps are the first requirements of advanced agriculture app lications such as yield forecasting . Early - season crop yield estimates are a crucial factor for food security and monitor ing agricultural subventio ns. Crop maps are also an essential tool for statistical purposes to analyze annual changes in agricultural p roduction. However, there are a variety of field crops with similar phenologies and spectral signatures.

Distributed machine learning training is one of the most common and important workloads running on data centers today, but it is rarely executed alone. Instead, to reduce costs, computing resources are consolidated and shared by different applications. In this scenario, elasticity and proper load balancing are vital to maximize efficiency, fairness, and utilization. Currently, most distributed training frameworks do not support the aforementioned properties. A few exceptions that do support elasticity, imitate generic distributed frameworks and use micro-tasks. In this paper we illustrate that micro-tasks are problematic for machine learning applications, because they require a high degree of parallelism which hinders the convergence of distributed training at a pure algorithmic level (i.e., ignoring overheads and scalability limitations). To address this, we propose Chicle, a new elastic distributed training framework which exploits the nature of machine learning algorithms to implement elasticity and load balancing without micro-tasks. We use Chicle to train deep neural network as well as generalized linear models, and show that Chicle achieves performance competitive with state of the art rigid frameworks, while efficiently enabling elastic execution and dynamic load balancing.

Automated medical prognosis has gained interest as artificial intelligence evolves and the potential for computer-aided medicine becomes evident. Nevertheless, it is challenging to design an effective system that, given a patient's medical history, is able to predict probable future conditions. Previous works, mostly carried out over private datasets, have tackled the problem by using artificial neural network architectures that cannot deal with low-cardinality datasets, or by means of non-generalizable inference approaches. We introduce a Deep Learning architecture whose design results from an intensive experimental process. The final architecture is based on two parallel Minimal Gated Recurrent Unit networks working in bi-directional manner, which was extensively tested with the open-access Mimic-III dataset. Our results demonstrate significant improvements of automated medical prognosis, as measured with Recall@k. We summarize our experience as a set of relevant insights for the design of Deep Learning architectures. Our work improves the performance of computer-aided medicine and can serve as a guide in designing artificial neural networks used in prediction tasks.

Manifold assumption in learning states that: the data lie approximately on a manifold of much lower dimension than the input space. Generative models learn to generate data according to the underlying data distribution. Generative models are used in various tasks, such as data augmentation and generating variation of images. This paper addresses the following question: do generative models need to be aware of the topology of the underlying data manifold in which the data lie? This paper suggests that the answer is yes and demonstrates that these can have ramifications on security-critical applications, such as generative-model based defenses for adversarial examples. We provide theoretical and experimental results to support our claims.

Electroencephalography (EEG) classification techniques have been widely studied for human behavior and emotion recognition tasks. But it is still a challenging issue since the data may vary from subject to subject, may change over time for the same subject, and maybe heterogeneous. Recent years, increasing privacy-preserving demands poses new challenges to this task. The data heterogeneity, as well as the privacy constraint of the EEG data, is not concerned in previous studies. To fill this gap, in this paper, we propose a heterogeneous federated learning approach to train machine learning models over heterogeneous EEG data, while preserving the data privacy of each party. To verify the effectiveness of our approach, we conduct experiments on a real-world EEG dataset, consisting of heterogeneous data collected from diverse devices. Our approach achieves consistent performance improvement on every task.

Existing methods on visual emotion analysis mainly focus on coarse-grained emotion classification, i.e. assigning an image with a dominant discrete emotion category. However, these methods cannot well reflect the complexity and subtlety of emotions. In this paper, we study the fine-grained regression problem of visual emotions based on convolutional neural networks (CNNs). Specifically, we develop a Polarity-consistent Deep Attention Network (PDANet), a novel network architecture that integrates attention into a CNN with an emotion polarity constraint. First, we propose to incorporate both spatial and channel-wise attentions into a CNN for visual emotion regression, which jointly considers the local spatial connectivity patterns along each channel and the interdependency between different channels. Second, we design a novel regression loss, i.e. polarity-consistent regression (PCR) loss, based on the weakly supervised emotion polarity to guide the attention generation. By optimizing the PCR loss, PDANet can generate a polarity preserved attention map and thus improve the emotion regression performance. Extensive experiments are conducted on the IAPS, NAPS, and EMOTIC datasets, and the results demonstrate that the proposed PDANet outperforms the state-of-the-art approaches by a large margin for fine-grained visual emotion regression. Our source code is released at: https://github.com/ZizhouJia/PDANet.

Multiple Instance learning (MIL) algorithms are tasked with learning how to associate sets of elements with specific set-level outputs. Towards this goal, the main challenge of MIL lies in modelling the underlying structure that characterizes sets of elements. Existing methods addressing MIL problems are usually tailored to address either: a specific underlying set structure; specific prediction tasks, e.g. classification, regression; or a combination of both. Here we present an approach where a set representation is learned, iteratively, by looking at the constituent elements of each set one at a time. The iterative analysis of set elements enables our approach with the capability to update the set representation so that it reflects whether relevant elements have been detected and whether the underlying structure has been matched. These features provide our method with some model explanation capabilities. Despite its simplicity, the proposed approach not only effectively models different types of underlying set structures, but it is also capable of handling both classification and regression tasks - all this while requiring minimal modifications. An extensive empirical evaluation shows that the proposed method is able to reach and surpass the state-of-the-art.

Automated in - vitro cell detection and counting ha ve been a key theme for artificial and intelligent biological analysis such as biopsy, drug analysis and decease diagnosis. Along with the rapid development of microfluidics and lab - on - chip technolog ies, in - vitro live cell analysis has be en one of the critical task s for both research and industry communities. However, it is a great challenge to obtain and then predict the precis e information of liv e cells from numerous microscopic videos and images. In this paper, we investigated in - vitro detection of white blood cell s using deep neural networks, and discuss ed how state - of - the - art machine learning techniques could fulfil the needs of medical diagnos is. The approach we used in this study wa s based on Faster Region - based Convolutional Neural Networks (Faster RCNNs), and a transfer learning process wa s applied to apply this technique to the microscopic detection of blood cell s . Our experimental results demonstrated that fast and efficient analysis of blood cell s via automated microscopic imaging can achieve much better accuracy and faster speed than the conventionally applied methods, implying a promising future of this technology to be applied to the microfluidic point - of - care medical devices .