Reading's research outputs online Title：Portfolio formation with preselection using deep learning from long-term financial data

The supervision of P2P lending platforms has always been a hot topic. However, if the government regulates the platform too strictly, it would restrain the subjective initiative of Internet financial innovation, and if the government overstimulates the platform, it might lead to systemic financial risks. From the perspective of government double objective optimization, this article sets a specific scene and analyzes the policy choice of supervision and incentive of P2P platform through game theory modeling and numerical simulation. Two schemes are offered, respectively, "First regulate and then motivate" and "First motivate and then regulate". The results show that the government should first motivate and then regulate the P2P lending platforms, so as to achieve the government's dual objective optimization and utility maximization. Moreover, the investment of supervision and incentive should be adjusted continuously with the development of the industry.

Purpose The purpose of this paper is to construct a multi-relational network for an online sharing platform in the age of the sharing economy, to identify the factors impacting users’ product adoption behavior and to predict consumers’ purchases of user-generated products on the platform. Design/methodology/approach The study conducted multi-relational network analyses of five different sub-networks in identifying influential factors for e-book adoption. Meanwhile, the study adopted machine learning methods with different classification algorithms and feature sets to predict users’ purchasing behaviors. Findings The authors found that an individual’s adoption of a product was correlated with his or her purchasing habits and collaboration with others on the online sharing platform. Through the inclusion of network features, the authors were able to build a predictive model that forecasted consumers’ purchases of user-generated e-books with reasonable accuracy. Research limitations/implications The interdisciplinary approach used in the study can serve as a good reference for identifying factors impacting the product adoption behavior of users in the online sharing platform, through employing different sociological and computational methods. Practical implications The outcome of the study has provided important managerial implications, especially for the design of social commerce platform in the age of the sharing economy. Social implications The authors verified the social influence impacting consumers’ product adoption behavior and shed light on the value of collaboration in the age of the sharing economy. Originality/value The study was the first to identify user-generated e-book adoption on an online sharing platform from a multi-relational network perspective. The idea and the approach supplied a new method of behavioral analysis in the context of a sharing economy.

No abstract

Based on Hartmann-Shack sensor technique, an online thin film stress measuring system was introduced to measure the film stresses of TiO 2 and SiO 2 , and comparison was made between the film stresses prepared respectively by the conventional process and the ion-beam assisted deposition. The effect of ion-beam assisted deposition on the film stresses of TiO 2 and SiO 2 was investigated in details, and the stress control methodologies using on-line adjustment and film doping were put forward. The results show that the film stress value of TiO 2 prepared by ion-beam assisted deposition is 40 MPa lower than that prepared by conventional process, and the stress of TiO 2 film changes gradually from tensile stress into compressive stress with increasing ion energy; while the film stress of SiO 2 is a tensile stress under ion-beam assisted deposition because of the ion-beam sputtering effect, and the film refractive index decreases with increasing ion energy. A dynamic film stress control can be achieved through in-situ adjustment of the processing parameters based on the online film stress measuring technique, and the intrinsic stress of film can be effectively changed through film doping.