Stock price prediction is an important issue in the financial world, as it contributes to the development of effective strategies for stock exchange transactions. In this paper, we propose a generic framework employing Long Short-Term Memory (LSTM) and convolutional neural network (CNN) for adversarial training to forecast high-frequency stock market. This model takes the publicly available index provided by trading software as input to avoid complex financial theory research and difficult technical analysis, which provides the convenience for the ordinary trader of nonfinancial specialty. Our study simulates the trading mode of the actual trader and uses the method of rolling partition training set and testing set to analyze the effect of the model update cycle on the prediction performance. Extensive experiments show that our proposed approach can effectively improve stock price direction prediction accuracy and reduce forecast error.

Identifying influential nodes is an important topic in many diverse applications, such as accelerating information propagation, controlling rumors and diseases. Many methods have been put forward to identify influential nodes in complex networks, ranging from node centrality to diffusion-based processes. However, most of the previous studies do not take into account overlapping communities in networks. In this paper, we propose an effective method based on network representation learning. The method considers not only the overlapping communities in networks, but also the network structure. Experiments on real-world networks show that the proposed method outperforms many benchmark algorithms and can be used in large-scale networks.

Large-scale terminals’ various QoS requirements are key challenges confronting the resource allocation of Satellite Internet of Things (S-IoT). This paper presents a deep reinforcement learning-based online channel allocation and power control algorithm in an S-IoT uplink scenario. The intelligent agent determines the transmission channel and power simultaneously based on contextual information. Furthermore, the weighted normalized reward concerning success rate, power efficiency, and QoS requirement is adopted to balance the performance between increasing resource efficiency and meeting QoS requirements. Finally, a practical deployment mechanism based on transfer learning is proposed to promote onboard training efficiency and to reduce computation consumption of the training process. The simulation demonstrates that the proposed method can balance the success rate and power efficiency with QoS requirement guaranteed. For S-IoT’s normal operation condition, the proposed method can improve the power efficiency by 60.91% and 144.44% compared with GA and DRL_RA, while its power efficiency is only 4.55% lower than that of DRL-EERA. In addition, this method can be transferred and deployed to a space environment by merely 100 onboard training steps.

MEC (Mobile Edge Computing) provides both IT service environment and cloud computation on the edge of the network. This technology not only minimizes the end-to-end latency but also increases the efficiency of computing. Some latency-sensitive applications, such as cloud video, online game, and augmented reality, take advantage of the MEC system to provide fast and stable services. Several new network techniques, including the implementation of NFV (Network Function Virtualization), the placement of VNF (Virtual Network Function) and the scheduling of SFC (Service Function Chain), should be considered to be applied in the MEC system. In this paper, we focus on the research about the scheduling of SFC in the NFV enabled MEC system and propose a solution accordingly. First, we make reasonable assumptions on the settings of MEC systems and model the SFC scheduling problem into a flexible job-shop scheduling problem. Since minimizing the latency can significantly improve the quality of service (QoS) and increase the revenue of Internet Service Providers, our optimization goal is to minimize the overall scheduling latency. To solve this optimization problem, a deep reinforcement learning based algorithm DQS is proposed. DQS can detect the variation of the MEC system’s environment and perform adaptive scheduling for SFC requests. As the results of the simulation indicate, DQS works better than the other off-the-shelf algorithms in two key indexes: overall scheduling latency and average resource usage. Moreover, DQS can shorten the decision time and schedule SFCs stably with high performance. It is suitable to be extended to an online scheduling algorithm.

Background: Previously, microRNA (miR)-7 has been reported to function as a tumor suppressor in human cancers, but the correlations of miR-7 expression with prognosis and cisplatin (CDDP) resistance in lung adenocarcinoma (LA) are unclear. Here, our aim is to determine the prognostic significance of miR-7 and its roles in the regulation of CDDP resistance in LA. Methods: Quantitative real-time PCR (qRT-PCR) assay was performed to determine miR-7 expression in 108 paired of LA tissues and analyze its correlations with clinicopathological factors of patients. The patient survival data were collected retrospectively by Kaplan-Meier analyses, and multivariate analysis was performed using the Cox proportional hazards model to determine the prognostic significance of miR-7 expression. The effects of miR-7 expression on the chemosensitivity of LA cells to CDDP and its possible mechanisms were evaluated by MTT, flow cytometry, Western blot and luciferase assays. Results: It was observed that the relative expression level of miR-7 in LA tissues was significantly lower than that in the adjacent normal tissues and low miR-7 expression level was closely associated with poorer tumor differentiation, advanced pathological T-factor, higher incidence of lymph node metastasis and advanced p-TNM stage. Also, patients with low miR-7 expression showed a shorter overall survival than those with high miR-7 expression, and multivariate analysis indicated that status of miR-7 expression was an independent molecular biomarker for predicting the overall survival (OS) of LA patients. In addition, upregulation of miR-7 increases the sensitivity of LA cells to CDDP via induction of apoptosis by targeting Bcl-2. Conclusions: Our finding for the first time demonstrates that low miR-7 expression may be an independent poor prognostic factor and targeting miR-7 may be a potential strategy for the reversal of CDDP resistance in LA.