Shandong University

About: Shandong University is a education organization based out in Jinan, Shandong, China. It is known for research contribution in the topics: Laser & Cancer. The organization has 99070 authors who have published 99160 publications receiving 1625094 citations. The organization is also known as: Shāndōng Dàxué.

Improvement of Generalization Ability of Deep CNN via Implicit Regularization in Two-Stage Training Process

Abstract: Optimization of deep learning is no longer an imminent problem, due to various gradient descent methods and the improvements of network structure, including activation functions, the connectivity style, and so on. Then the actual application depends on the generalization ability, which determines whether a network is effective. Regularization is an efficient way to improve the generalization ability of deep CNN, because it makes it possible to train more complex models while maintaining a lower overfitting. In this paper, we propose to optimize the feature boundary of deep CNN through a two-stage training method (pre-training process and implicit regularization training process) to reduce the overfitting problem. In the pre-training stage, we train a network model to extract the image representation for anomaly detection. In the implicit regularization training stage, we re-train the network based on the anomaly detection results to regularize the feature boundary and make it converge in the proper position. Experimental results on five image classification benchmarks show that the two-stage training method achieves a state-of-the-art performance and that it, in conjunction with more complicated anomaly detection algorithm, obtains better results. Finally, we use a variety of strategies to explore and analyze how implicit regularization plays a role in the two-stage training process. Furthermore, we explain how implicit regularization can be interpreted as data augmentation and model ensemble.

A Metal−Organic Nanotube Exhibiting Reversible Adsorption of (H2O)12 Cluster

Abstract: A novel 1D metal−organic nanotube, exhibiting reversible and fast adsorption of the (H2O)12 cluster, has been synthesized and characterized.

Trust prediction and trust-based source routing in mobile ad hoc networks

Abstract: Mobile ad hoc networks (MANETs) are spontaneously deployed over a geographically limited area without well-established infrastructure. The networks work well only if the mobile nodes are trusty and behave cooperatively. Due to the openness in network topology and absence of a centralized administration in management, MANETs are very vulnerable to various attacks from malicious nodes. In order to reduce the hazards from such nodes and enhance the security of network, this paper presents a dynamic trust prediction model to evaluate the trustworthiness of nodes, which is based on the nodes’ historical behaviors, as well as the future behaviors via extended fuzzy logic rules prediction. We have also integrated the proposed trust predication model into the Source Routing Mechanism. Our novel on-demand trust-based unicast routing protocol for MANETs, termed as Trust-based Source Routing protocol (TSR), provides a flexible and feasible approach to choose the shortest route that meets the security requirement of data packets transmission. Extensive experiments have been conducted to evaluate the efficiency and effectiveness of the proposed mechanism in malicious node identification and attack resistance. The results show that TSR improves packet delivery ratio and reduces average end-to-end latency.

RepeatNet: A Repeat Aware Neural Recommendation Machine for Session-Based Recommendation

Abstract: Recurrent neural networks for session-based recommendation have attracted a lot of attention recently because of their promising performance. repeat consumption is a common phenomenon in many recommendation scenarios (e.g., e-commerce, music, and TV program recommendations), where the same item is re-consumed repeatedly over time. However, no previous studies have emphasized repeat consumption with neural networks. An effective neural approach is needed to decide when to perform repeat recommendation. In this paper, we incorporate a repeat-explore mechanism into neural networks and propose a new model, called RepeatNet, with an encoder-decoder structure. RepeatNet integrates a regular neural recommendation approach in the decoder with a new repeat recommendation mechanism that can choose items from a user’s history and recommends them at the right time. We report on extensive experiments on three benchmark datasets. RepeatNet outperforms state-of-the-art baselines on all three datasets in terms of MRR and Recall. Furthermore, as the dataset size and the repeat ratio increase, the improvements of RepeatNet over the baselines also increase, which demonstrates its advantage in handling repeat recommendation scenarios.