This paper proposes an approach for applying GANs to NMT. We build a conditional sequence generative adversarial net which comprises of two adversarial sub models, a generator and a discriminator. The generator aims to generate sentences which are hard to be discriminated from human-translated sentences ( i.e., the golden target sentences); And the discriminator makes efforts to discriminate the machine-generated sentences from humantranslated ones. The two sub models play a mini-max game and achieve the win-win situation when they reach a Nash Equilibrium. Additionally, the static sentence-level BLEU is utilized as the reinforced objective for the generator, which biases the generation towards high BLEU points. During training, both the dynamic discriminator and the static BLEU objective are employed to evaluate the generated sentences and feedback the evaluations to guide the learning of the generator. Experimental results show that the proposed model consistently outperforms the traditional RNNSearch and the newly emerged state-ofthe-art Transformer on English-German and Chinese-English translation tasks.

Unsupervised neural machine translation (NMT) is a recently proposed approach for machine translation which aims to train the model without using any labeled data. The models proposed for unsupervised NMT often use only one shared encoder to map the pairs of sentences from different languages to a shared-latent space, which is weak in keeping the unique and internal characteristics of each language, such as the style, terminology, and sentence structure. To address this issue, we introduce an extension by utilizing two independent encoders but sharing some partial weights which are responsible for extracting high-level representations of the input sentences. Besides, two different generative adversarial networks (GANs), namely the local GAN and global GAN, are proposed to enhance the cross-language translation. With this new approach, we achieve significant improvements on English-German, English-French and Chinese-to-English translation tasks.

Aim: Physcion is a major bioactive ingredient in the traditional Chinese medicine Radix et Rhizoma Rhei, which has an anthraquinone chemical structure and exhibits a variety of pharmacological activities including laxative, hepatoprotective, anti-inflammatory, antimicrobial and anti-proliferative effects. In this study we investigated the effect of physcion on human nasopharyngeal carcinoma in vitro and in vivo, as well as the mechanisms underlying the anti-tumor action. Methods: The nasopharyngeal carcinoma cell line CNE2 was treated with physcion, and cell viability was detected using MTT and colony formation assays. Flow cytometry was used to assess the cell cycle arrest, mitochondrial membrane potential loss, apoptosis, autophagy and intracellular ROS generation. Apoptotic cell death was also confirmed by a TUNEL assay. The expression of target or marker molecules was determined using Western blotting. The activity of caspase-3, 8, and 9 was detected with an ELISA kit. A xenograft murine model was used to evaluate the in vivo anti-tumor action of physcion, the mice were administered physcion (10, 20 mg·kg -1 ·d -1 , ip) for 30 d.Results: Treatment with physcion (5, 10, and 20 μmol/L) dose-dependently suppressed the cell viability and colony formation in CNE2 cells. Physcion (10 and 20 μmol/L) dose-dependently blocked cell cycle progression at G 1 phase and induced both caspase-dependent apoptosis and autophagy in CNE2 cells. Furthermore, physcion treatment induced excessive ROS generation in CNE2 cells, and subsequently disrupted the miR-27a/ZBTB10 axis, resulting in repression of the transcription factor Sp1 that was involved in physcioninduced apoptosis and autophagy. Moreover, physcion-induced autophagy acted as a pro-apoptotic factor, and possibly contributed to physcion-induced apoptosis. In the xenograft murine model, administration of physcion dose-dependently suppressed the tumor growth without affecting the body weight. Furthermore, the anti-tumor effects of physcion were correlated with downregulation of Sp1 and suppression of miR-27a in the tumor tissues. Conclusion: Physcion induces apoptosis and autophagy in human nasopharyngeal carcinoma by targeting Sp1, which was mediated by ROS/miR-27a/ZBTB10 signaling. The results suggest that physcion is a promising candidate for the treatment of human nasopharyngeal carcinoma.

No abstract

This paper proposes a new pre-training method, called Code-Switching Pre-training (CSP for short) for Neural Machine Translation (NMT). Unlike traditional pre-training method which randomly masks some fragments of the input sentence, the proposed CSP randomly replaces some words in the source sentence with their translation words in the target language. Specifically, we firstly perform lexicon induction with unsupervised word embedding mapping between the source and target languages, and then randomly replace some words in the input sentence with their translation words according to the extracted translation lexicons. CSP adopts the encoderdecoder framework: its encoder takes the codemixed sentence as input, and its decoder predicts the replaced fragment of the input sentence. In this way, CSP is able to pre-train the NMT model by explicitly making the most of the cross-lingual alignment information extracted from the source and target monolingual corpus. Additionally, we relieve the pretrainfinetune discrepancy caused by the artificial symbols like [mask]. To verify the effectiveness of the proposed method, we conduct extensive experiments on unsupervised and supervised NMT. Experimental results show that CSP achieves significant improvements over baselines without pre-training or with other pre-training methods.

With the development of unmanned aerial vehicle (UAV) and artificial intelligence (AI) technology, Intelligent UAV will be widely used in future autonomous aerial combat. Previous researches on autonomous aerial combat within visual range (WVR) have limitations due to simplifying assumptions, limited robustness, and ignoring sensor errors. In this paper, in order to consider the error of the aircraft sensors, we model the aerial combat WVR as a state-adversarial Markov decision process (SA-MDP), which introduce the small adversarial perturbations on state observations and these perturbations do not alter the environment directly, but can mislead the agent into making suboptimal decisions. Meanwhile, we propose a novel autonomous aerial combat maneuver strategy generation algorithm with high-performance and high-robustness based on state-adversarial deep deterministic policy gradient algorithm (SA-DDPG), which add a robustness regularizers related to an upper bound on performance loss at the actor-network. At the same time, a reward shaping method based on maximum entropy (MaxEnt) inverse reinforcement learning algorithm (IRL) is proposed to improve the aerial combat strategy generation algorithm’s efficiency. Finally, the efficiency of the aerial combat strategy generation algorithm and the performance and robustness of the resulting aerial combat strategy is verified by simulation experiments. Our main contributions are three-fold. First, to introduce the observation errors of UAV, we are modeling air combat as SA-MDP. Second, to make the strategy network of air combat maneuver more robust in the presence of observation errors, we introduce regularizers into the policy gradient. Third, to solve the problem that air combat’s reward function is too sparse, we use MaxEnt IRL to design a shaping reward to accelerate the convergence of SA-DDPG.

SUMMARYThis paper presents an experimental study, while a companion paper addresses an analytical study, to explore the possibility of using a hybrid platform to mitigate vibration of a batch of high-tech equipment installed in a building subject to nearby tra c-induced ground motion. A three-storey building model and a hybrid platform model are designed and manufactured. The hybrid platform is mounted on the building oor through passive mounts composed of leaf springs and oil dampers and controlled actively by an electromagnetic actuator with velocity feedback control strategy. The passive mounts are designed in such a way that the sti ness and damping ratio of the platform can be changed. A series of shaking table tests are then performed on the building model without the platform, with the passive platform of di erent parameters, and with the hybrid platform. The experimental results demonstrate that the hybrid platform is very e ective in reducing the velocity response of a batch of high-tech equipment in the building subject to nearby tra c-induced ground motion if dynamic properties of the platform and control feedback gain are selected appropriately.

No abstract