Stock price prediction and modeling demonstrate high economic value in the financial market. Due to the non-linearity and volatility of stock prices and the unique nature of financial transactions, it is essential for the prediction method to ensure high prediction performance and interpretability. However, existing methods fail to achieve both the two goals simultaneously. To fill this gap, this paper presents an interpretable intuitionistic fuzzy inference model, dubbed as IIFI. While retaining the prediction accuracy, the interpretable module in IIFI can automatically calculate the feature contribution based on the intuitionistic fuzzy set, which provides high interpretability of the model. Also, most of the existing training algorithms, such as LightGBM, XGBoost, DNN, Stacking, etc, can be embedded in the inference module of our proposed model and achieve better prediction results. The back-test experiment on China’s A-share market shows that IIFI achieves superior performance — the stock profitability can be increased by more than 20% over the baseline methods. Meanwhile, interpretable results show that IIFI can effectively distinguish between important and redundant features via rating corresponding scores to each feature. As a byproduct of our interpretable methods, the scores over features can be used to further optimize the investment strategy.

Repair and maintenance services are among the most lucrative aspects of the entire automobile business chain. However, in the context of fierce competition, customer churns have led to the bankruptcy of several 4S (sales, spare parts, services, and surveys) shops. In this regard, a six-year dataset is utilized to study customer behaviors to aid managers identify and retain valuable but potential customer churn through a customized retention solution. First, we define the absence and presence behaviors of customers and thereafter generate absence data according to customer habits; this makes it possible to treat the customer absence prediction problem as a classification problem. Second, the repeated absence and presence behaviors of customers are considered as a whole from a lifecycle perspective. A modified recurrent neural network (RNN-2L) is proposed; it is more efficient and reasonable in structure compared with traditional RNN. The time-invariant customer features and the sequential lifecycle features are handled separately; this provides a more sensible specification of the RNN structure from a behavioral interpretation perspective. Third, a customized retention solution is proposed. By comparing the proposed model with those that are conventional, it is found that the former outperforms the latter in terms of area under the curve (AUC), confusion matrix, and amount of time consumed. The proposed customized retention solution can achieve significant profit increase. This paper not only elucidates the customer relationship management in the automobile aftermarket (where the absence and presence behaviors are infrequently considered), but also presents an efficient solution to increase the predictive power of conventional machine learning models. The latter is achieved by considering behavioral and business perspectives.

This paper proposes a trustable mold redesign knowledge-sharing platform, known as CKshare, based on private cloud and blockchain technology. Firstly, we use private cloud to store the mold redesign knowledge of each party to meet its own privacy and data format requirements. Secondly, a blockchain network is used for recording the knowledge and its transactions to ensure security and trustfulness. Thirdly, a simple retrieval mechanism is developed based on k-nearest neighbors for retrieving the codified knowledge on the platform. To realize CKshare, a prototype platform has been developed and explained based on real data from the case mold company.

基于SDN与NFV的服务部署是提升网络性能与网络服务的重要技术,但是在服务部署过程中会出现一些安全问题,如服务链被篡改、截取、重放等.利用拟态防御理论,提出一种基于拟态防御的SDN服务部署架构,该架构利用动态异构冗余的特点,对服务部署过程中的主要攻击实现主动防御,有效增加攻击难度.结合服务部署的特点,改进了拟态防御中的调度机制和判决机制,增加了系统的异构程度,提升了系统的安全性与执行效率.

With the rapid development of network technologies, many new technologies, such as Software Defined Network (SDN), are applied to firewalls to manage network security. However, current SDN firewalls cannot automatically change security policies according to dynamic network status or deploy personalized policy based on user identities. In this paper, we design a special SDN switch that incorporates a traffic acquisition module and a data analysis module. According to the traffic patterns caused by different user behaviors, the proposed switch could recognize the user identities by statistical analysis and clustering analysis, and automatically deploy corresponding network policies. Experiments conducted over an OpenvSwitch showed the proposed SDN switch could accurately identify three kinds of users and apply respective flow tables successfully.

Software defined network (SDN) is a new kind of network technology,and the security problems are the hot topics in SDN field,such as SDN control channel security,forged service deployment and external distributed denial of service (DDoS) attacks.Aiming at DDoS attack problem of security in SDN,a DDoS attack detection method called DCNN-DSAE based on deep learning hybrid model in SDN was proposed.In this method,when a deep learning model was constructed,the input feature included 21 different types of fields extracted from the data plane and 5 extra self-designed features of distinguishing flow types.The experimental results show that the method has high accuracy,it’s better than the traditional support vector machine (SVM) and deep neural network (DNN) and other machine learning methods.At the same time,the proposed method can also shorten the processing time of classification detection.The detection model is deployed in SDN controller,and the new security policy is sent to the OpenFlow switch to achieve the defense against specific DDoS attack.

A network resource pricing strategy based on SDN framework was proposed to solve the problem of resource overload and network congestion caused by resource requests in the network.Firstly,the user’s needs and consumption patterns were analyzed,and the resource transaction model was determined.Secondly,combined with the spot consumption model and plan consumption,the reservation mechanism under the resource dynamic pricing strategy model was determined.Finally,according to the proposed resource pricing strategy,the resource price was simulated,the results show that by stimulating the user to reserve resources in advance and combining the trust degree and consumption habits of each user with corresponding preferential prices,it is possible to effectively avoid the users from obtaining network resources in a centralized manner.

In OpenFlow-based SDN(software defined network),applications can be deployed through dispatching the flow polices to the switches by the application orchestrator or controller.Policy conflict between multiple applications will affect the actual forwarding behavior and the security of the SDN.With the expansion of network scale of SDN and the increasement of application number,the number of flow entries will increase explosively.In this case,traditional algorithms of conflict detection will consume huge system resources in computing.An intelligent conflict detection approach based on deep learning was proposed which proved to be efficient in flow entries' conflict detection.The experimental results show that the AUC (area under the curve) of the first level deep learning model can reach 97.04％,and the AUC of the second level model can reach 99.97％.Meanwhile,the time of conflict detection and the scale of the flow table have a linear growth relationship.

Performance assessment of technology-based ventures requires consideration of the nature of their businesses and the dynamics of their emerging industries. This paper explores the development of a cluster-based and quantitative measurement system for science and technology parks to evaluate the performance of technology-based ventures. The proposed method incorporates technique for order preference by similarity to ideal solution (TOPSIS) and weight allocation. It ranks the technology-based ventures in different technological clusters, based on a range of indicators pertinent to productivity, research and development (R&D) effort, R&D personnel percentage, time to market and financial performance. This method has been implemented through a trial study conducted within the Hong Kong Science and Technology Parks Corporation. The results indicate that R&D spending has a strong impact on a company's performance ranking. The performance of technology-based ventures should be measured with respect to their R&D investments and their pertinent efforts to commercialise products.

As a new type of network architecture, software defined network (SDN) breaks the closed mode of traditional network, realizes the separation between forwarding and control, and therefore can effectively solve a series of problems of the current Internet. Taking advantage of economic principles and according to the characteristics of the traditional auction model, we propose a price negotiation algorithm, namely multi-homing combinatorial double auction model (MCDAM) algorithm, to curb the waste of SDN resources. MCDAM can solve the problem of reasonable transaction price between the buyers and sellers, make the price as a lever to achieve the purpose of SDN resources reasonable allocation. Simulation results illustrate the superiority and feasibility of the proposed algorithm.

SummaryThrough network programmability, software defined network can simplify network control and management. Since the current software defined network southbound interface level is low and programming situation is complex, it requires a high‐level abstract programming language to simplify programming. First, this paper improves the NetCore programming language to generate NetCore‐M language, so that it can support deployment of multipolicies combination including packet drop action. This paper describes in detail the syntax, semanteme, and implementation of NetCore‐M language forwarding policy service. Secondly, this paper describes the network policy conflict systematically. Finally, this paper shows that the modified multipolicies combination algorithm can effectively detect policies conflicts based on the implementation of the Pyretic project.

Distributed denial of service (DDoS) is a special form of denial of service (DoS) attack based on denial of service(DoS).It is a distributed,collaborative large-scale network attack.A DDoS detection method based on deep learning was presented.The method included two stages:feature processing and model detection:feature extraction,format conversion and dimension reconstruction of the input data packet was performed in feature processing stage;in the model detection stage,the processed features were input to the depth learning network model to detect whether the input data packets was DDoS attack packet.The model was trained by the ISCX2012 dataset,and the model was validated by real-time DDoS attack.The experimental results show that DDoS attack detection method based on deep learning has high detection precision,little dependency on hardware and software equipment,and the model of depth learning network is easy to update.

Centralized network control plane in SDN brings scalability and reliability problem to the network, therefore, the research of multi-controller is appeared. For improving the communication efficiency between the controller and the network device, this paper proposes a loose management strategy to dynamically adjust the frequency of interaction between controllers and network devices. Based on the above idea, firstly, this paper designed the scheme and algorithm of multi-controller loose management. Secondly, this paper quantitatively analyzed the advantages of multi-controller loose management algorithm by mathematically modeling the virtual network deployment success ratio and the management revenue between controllers and network devices. Finally, experiment results show that the multi-controller loose management idea can improve the communication efficiency between the controller and the network device and the controller management efficiency. Simulation results also show that mathematical model accurately predict the performance of loose management algorithm.

The traditional network architecture cannot meet the growing needs of increasingly diverse network applications. The new generation network architectures, software defined network (SDN) which have features of high openness, flexibility, scalability and high controllability have widely been studied. However, a series of new problems are also produced in this new architecture, such as security problem resulted from the high openness, and performance problem resulted from the high flexibility. Currently, the researchers are mainly concentrated in the implementation and standardisation of SDN. There is no good approach to evaluate and optimise the system performance. This studies the performance problems unresolved in SDN research area, and proposed an idea and method to analyse the control channel of ForCES (Forwarding and Control Element Separation)-based SDN by using the stochastic network calculus theory. First, the architecture of ForCES-based SDN is introduced. Then, the performance model of the channel is given. Based on the model, an optimisation method to performance is detailed. This study also gives a simulation by using NS-2 (Network Simulator, Version 2) to verify the correctness and reliability of the performance model.

Micro service is a new method of modern application architecture,its characteristic is that decomposing the complex applications into discrete,independent service,so that decoupling of applications scheme can be achieved,and the individual service could be developed,deployed and extensioned independently.The change of the application must drive on the change of the underlying network,the architecture of reconfigurable fundamental information communication network aimd to compensate for the network business requirements and network transmission capacity through the network reconfiguration mechanism.A meta model construction mechanism with the SDN architecture for fusion the reconfiguration and software defined network technology was proposed,it aimed to present the services of each layer in the SDN architecture by the form of meta model.And after referencing the AKF scale cube,a kind of meta model building method based on orthogonal decomposition was proposed.Finally,the feasibility of meta model building method based on orthogonal decomposition from the theoretical analysis and development of example was verified.

SDN network as new network architecture, can address a range of issues of the current internet effectively, this paper in the view of SDN network resources' allocation, for decreasing resources waste and achieving a reasonable distribution of SDN resources, come up with a suitable resource price negotiation algorithm based on trust mechanism in SDN network environment. According to the characteristics of the trust mechanism, simulating on the unit utility and equivalent price of SDN network resources in three aspects (remove individual malicious nodes, remove most malicious nodes and remove all malicious nodes), through the above simulation, combined trust mechanism with MCDAM algorithm, can form a weighing system which pricing and trust comprehensive competitive, so that the whole SDN trading environment become safer and more stable. So trust mechanism as the patch of SDN pricing negotiation algorithm. Finally, propose the planned trading model based on trust mechanism as a management method of SDN trading.

To address the issues that middleboxes as a fundamental part of today's networks are facing, Network Function Virtualization (NFV) has been recently proposed, which in essence asserts to migrate hardware-based middleboxes into software-based virtualized function entities. Due to the demands of virtual services placement in NFV network environment, this paper models the service amount placement problem involving with the resources allocation as a cooperative game and proposes the placement policy by Nash Bargaining Solution (NBS). Specifically, we first introduce the system overview and apply the rigorous cooperative game-theoretic guide to build the mathematical model, which can give consideration to both the responding efficiency of service requirements and the allocation fairness. Then a distributed algorithm corresponding to NBS is designed to achieve predictable network performance for virtual instances placement. Finally, with simulations under various scenarios, the results show that our placement approach can achieve high utilization of network through the analysis of evaluation metrics namely the satisfaction degree and fairness index. With the suitable demand amount of services, the average values of two metrics can reach above 90%. And by tuning the base placement, our solution can enable operators to flexibly balance the tradeoff between satisfaction and fairness of resources sharing in service platforms.

In recent years, due to the increase of network use rs, network resources wasting on a global scale is becoming more serious. Traditional resource scheduling policy can not meet the current allocation of network resourc es; we propose a SDN resource scheduling strategy based on bilater al market three-tier pricing mechanism. Respectivel y, each layer of the three-tier pricing mechanism integrates into the three-tier architecture of the SDN, and introd uces the concept of a bilateral market, changing SDN resource schedu ling into a commodity trading in economics.

This document discusses Control Element (CE) High Availability (HA) within a Forwarding and Control Element Separation (ForCES) Network Element (NE).Additionally, this document updates RFC 5810 by providing new normative text for the Cold Standby High Availability mechanism.