The present paper proposes generalized Gaussian kernel adaptive filtering, where the kernel parameters are adaptive and data-driven. The Gaussian kernel is parametrized by a center vector and a symmetric positive definite (SPD) precision matrix, which is regarded as a generalization of the scalar width parameter. These parameters are adaptively updated on the basis of a proposed least-square-type rule to minimize the estimation error. The main contribution of this paper is to establish update rules for precision matrices on the SPD manifold in order to keep their symmetric positive-definiteness. Different from conventional kernel adaptive filters, the proposed regressor is a superposition of Gaussian kernels with all different parameters, which makes such regressor more flexible. The kernel adaptive filtering algorithm is established together with a l1-regularized least squares to avoid overfitting and the increase of dimensionality of the dictionary. Experimental results confirm the validity of the proposed method.

Generative models (GMs) such as Generative Adversary Network (GAN) and Variational Auto-Encoder (VAE) have thrived these years and achieved high quality results in generating new samples. Especially in Computer Vision, GMs have been used in image inpainting, denoising and completion, which can be treated as the inference from observed pixels to corrupted pixels. However, images are hierarchically structured which are quite different from many real-world inference scenarios with non-hierarchical features. These inference scenarios contain heterogeneous stochastic variables and irregular mutual dependences. Traditionally they are modeled by Bayesian Network (BN). However, the learning and inference of BN model are NP-hard thus the number of stochastic variables in BN is highly constrained. In this paper, we adapt typical GMs to enable heterogeneous learning and inference in polynomial time.We also propose an extended autoregressive (EAR) model and an EAR with adversary loss (EARA) model and give theoretical results on their effectiveness. Experiments on several BN datasets show that our proposed EAR model achieves the best performance in most cases compared to other GMs. Except for black box analysis, we've also done a serial of experiments on Markov border inference of GMs for white box analysis and give theoretical results.

Drastic sudden large events in dynamical systems have become a key area of interest in a broad range of applications [2, 26]. From the perspective of modelling, a successful framework to capture many critical transitions has been to use systems with multiple time scales in combination with bifurcation theory [23]. The idea is that there are fast variables, which are driven slowly towards a bifurcation point, where the system can undergo a sudden jump for certain types of bifurcations. One aim in this context is to determine, whether there are early-warning signs for critical transitions, which can be computed from time series data before the actual event occurred. Groundbreaking work by Wiesenfeld in the 1980s [32] has already clearly shown that 1 precursors of bifurcations exist, and that they can be extracted from stochastic fluctuations based upon critical slowing down.

In this study, we propose the integration of competitive learning into convolutional neural networks (CNNs) to improve the representation learning and efficiency of fine-tuning. Conventional CNNs use back propagation learning, and it enables powerful representation learning by a discrimination task. However, it requires huge amount of labeled data, and acquisition of labeled data is much harder than that of unlabeled data. Thus, efficient use of unlabeled data is getting crucial for DNNs. To address the problem, we introduce unsupervised competitive learning into the convolutional layer, and utilize unlabeled data for effective representation learning. The results of validation experiments using a toy model demonstrated that strong representation learning effectively extracted bases of images into convolutional filters using unlabeled data, and accelerated the speed of the fine-tuning of subsequent supervised back propagation learning. The leverage was more apparent when the number of filters was sufficiently large, and, in such a case, the error rate steeply decreased in the initial phase of fine-tuning. Thus, the proposed method enlarged the number of filters in CNNs, and enabled a more detailed and generalized representation. It could provide a possibility of not only deep but broad neural networks.

Automatic generation of natural language from images has attracted extensive attention. In this paper, we take one step further to investigate generation of poetic language (with multiple lines) to an image for automatic poetry creation. This task involves multiple challenges, including discovering poetic clues from the image (e.g., hope from green), and generating poems to satisfy both relevance to the image and poeticness in language level. To solve the above challenges, we formulate the task of poem generation into two correlated sub-tasks by multi-adversarial training via policy gradient, through which the cross-modal relevance and poetic language style can be ensured. To extract poetic clues from images, we propose to learn a deep coupled visual-poetic embedding, in which the poetic representation from objects, sentiments and scenes in an image can be jointly learned. Two discriminative networks are further introduced to guide the poem generation, including a multi-modal discriminator and a poem-style discriminator. To facilitate the research, we have collected two poem datasets by human annotators with two distinct properties: 1) the first human annotated image-to-poem pair dataset (with 8,292 pairs in total), and 2) to-date the largest public English poem corpus dataset (with 92,265 different poems in total). Extensive experiments are conducted with 8K images generated with our model, among which 1.5K image are randomly picked for evaluation. Both objective and subjective evaluations show the superior performances against the state-of-art methods for poem generation from images. Turing test carried out with over 500 human subjects, among which 30 evaluators are poetry experts, demonstrates the effectiveness of our approach.

Traditional cross-modal retrieval assumes explicit association of concepts across modalities, where there is no ambiguity in how the concepts are linked to each other, e.g., when we do the image search with a query "dogs", we expect to see dog images. In this paper, we consider a different setting for cross-modal retrieval where data from different modalities are implicitly linked via concepts that must be inferred by high-level reasoning; we call this setting implicit concept association. To foster future research in this setting, we present a new dataset containing 47K pairs of animated GIFs and sentences crawled from the web, in which the GIFs depict physical or emotional reactions to the scenarios described in the text (called "reaction GIFs"). We report on a user study showing that, despite the presence of implicit concept association, humans are able to identify video-sentence pairs with matching concepts, suggesting the feasibility of our task. Furthermore, we propose a novel visual-semantic embedding network based on multiple instance learning. Unlike traditional approaches, we compute multiple embeddings from each modality, each representing different concepts, and measure their similarity by considering all possible combinations of visual-semantic embeddings in the framework of multiple instance learning. We evaluate our approach on two video-sentence datasets with explicit and implicit concept association and report competitive results compared to existing approaches on cross-modal retrieval.

The Fourth Industrial Revolution -- a global transitioning to a new set of systems and the integration of digital and physical technologies -- is upon us. Processing is rising exponentially, knowledge is becoming accessible to more people and information gathering is at an all-time high. Over the years, we've seen information technology have an increased role in the furtherance of humanity's cause for innovation and digital evolution. With new technologies such as artificial intelligence, the Internet of Things, energy storage and quantum computing, the future holds even greater potential for human development. The ubiquitous nature of information technology and its integration into virtually every facet of daily living has brought it within proximity of otherwise personal data.

The common conception of a technologically enabled apocalypse foresees a powerful artificial intelligence that, either deliberately or by accident, destroys human civilization. But as a new report from the RAND Corporation points out, the reality may be far subtler: As AI slowly erodes the foundations that made the Cold War possible, we may find ourselves hurtling towards all-out nuclear war. There's a "significant potential" for artificial intelligence to undermine the foundations of nuclear security, according to a new report published today by the RAND Corporation, a nonprofit, nonpartisan research organization. This grim conclusion was the product of a RAND workshop involving experts in AI, nuclear security, government, and military. The point of the workshop, which is part of RAND's Security 2040 project, was to evaluate the coming impacts of AI and advanced computing on nuclear security over the course of the next two decades.

Newswire) Autonomous Research, a global research firm focused on financial services, has announced the publication of Augmented Finance and Machine Intelligence, an in-depth look at the way artificial intelligence is transforming the financial services industry. Autonomous estimates that over $1 trillion of today's financial services cost structure could be replaced by machine learning and AI. This would affect 2.5 million employees in the US alone. This shift will impact operations across all functions and segments of the financial industry, from bank tellers to portfolio managers to insurance underwriters. By 2030, Autonomous expects $490 billion in costs to be exposed to AI in distribution, $350 billion in the middle office, and $200 billion in financial product manufacturing.

The American security agency CIA knows that the future can't go without artificial intelligence. The agency was all over the news last year because of Wikileaks which published their collection of hacking tools. CIA wants to deal with foreign spies, not human but AI-powered spies tracking CIA agents deployed overseas. An effective countermeasure would be using technology instead of humans to get the required intel. At a conference in Florida, CIA's Science and Technology division deputy director Dawn Meyerriecks talked about their AI developments without going into the details.