摘要：In view of huge search space in drug design, machine learning has become a powerful method to predict the affinity between small molecular drug and targeting protein with the development of artificial intelligence technology. However, various machine learning algorithms including massive different parameters make the prediction framework choice to be quite difficult. In this work, we took a recent drug design competition(from XtalPi company on the DataCastle platform) as the typical case to find the optimized parameters for different machines learning algorithms and the most effective algorithm. After the parameter optimizations, we compared the typical machine learning methods as decision tree(XGBoost, LightGBM) and artificial neural network(MLP, CNN) with root-mean-square error(RMSE) and coefficient of determination(R^2) evaluation. As a result, decision tree is more effective than the neural network as LightGBM>XGBoost>CNN>MLP in the affinity prediction of the specific drug design problem with ～160000 samples. For a much larger screening task in a more complicated drug design study, the sophisticated neural network model may go beyond the decision tree algorithm after generalization enhancing and overfitting reducing. The advanced machine learning methods could extract more information of protein-ligand bindings than traditional ones and improve the screen efficiency of drug design up to 200–1000 times.

摘要：In this paper,the MCI（multipath crossover interconnection）technique for octagon single and symmetrical spiral inductors has been presented to improve the quality *** metal wires of the single and symmetrical inductors formed by the top metal are divided into multiple segments according to the depth of the skin *** outermost path of the metal is crossover-interconnected to the innermost path by the underlayer metal and via The crossover technique makes the lengths of the total current paths between two ports approximately equal to each ***,the induced magnetic flux and resistance of each path can be balanced and the Q-factor of spiral inductors can be *** proposed MCI technique has been validated by the electromagnetic simulation with the 130-nm 1P6M SiGe BiCMOS *** the devices with occupying areas of 240 240 μm^2,results of electromagnetic simulation show that about 24%improvement in the Q-peak（3.3 GHz）of the MCI single inductor as compared to conventional single inductors（3.1 GHz）,and about 88.1%improvement in the Q-peak（3.2 GHz）of the MCI symmetrical inductor as compared to conventional symmetrical inductors（1.8 GHz）.

摘要：To describe the design approaches of IND-CCA2 （adaptive chosen ciphertext attack） secure public key encryption schemes systematically, the gaps between different kinds of intractable problems and IND-CCA2 security are studied. This paper points out that the construction of IND-CCA2 secure schemes is essentially to bridge these gaps. These gaps are categorized, analyzed and measured. Finally the methods to bridge these gaps are described. This explains the existing design approaches and gives an intuition about the difficulty of designing IND-CCA2 secure public key encryption schemes based on different types of assumptions.

摘要：In today＇s globalized digital world, networkbased, mobile, and interactive collaborations have enabled work platforms of personal computers to cross multiple geographical boundaries. The new requirements of privacy-preservation, sensitive information sharing, portability, remote attestation, and robust security create new problems in system design and implementation. There are critical demands for highly secure work platforms and security enhancing mechanisms for ensuring privacy protection, component integrity, sealed storage, and remote attestation of platforms. Trusted computing is a promising technology for enhancing the security of a platform using a trusted platform module （TPM）. TPM is a tamper-resistant microcontroller designed to provide robust security capabilities for computing platforms. It typically is affixed to the motherboard with a low pin count （LPC） bus. However, it limited in that TPM cannot be used directly in current common personal computers （PCs）, and TPM is not flexible and portable enough to be used in different platforms because of its interface with the PC and its certificate and key structure. For these reasons, we propose a portable trusted platform module （PTPM） scheme to build a trusted platform for the common PC based on a single cryptographic chip with a universal serial bus （USB） interface and extensible firmware interface （EFI）, by which platforms can get a similar degree of security protection in general-purpose systems. We show the structure of certificates and keys, which can bind to platforms via a PTPM and provide users with portability and flexibility in different platforms while still allowing the user and platform to be protected and attested. The implementation of prototype system is described in detail and the performance of the PTPM on cryptographic operations and time-costs of the system bootstrap are evaluated and analyzed. The results of experiments show that PTPM has high performances for supporting trusted computin

摘要：Head pose estimation has been considered an important and challenging task in computer vision. In this paper we propose a novel method to estimate head pose based on a deep convolutional neural network （DCNN） for 2D face images. We design an effective and simple method to roughly crop the face from the input image, maintaining the individual-relative facial features ratio. The method can be used in various poses. Then two convolutional neural networks are set up to train the head pose classifier and then compared with each other. The simpler one has six layers. It performs well on seven yaw poses but is somewhat unsatisfactory when mixed in two pitch poses. The other has eight layers and more pixels in input layers. It has better performance on more poses and more training samples. Before training the network, two reasonable strategies including shift and zoom are executed to prepare training samples. Finally, feature extraction filters are optimized together with the weight of the classification component through training, to minimize the classification error. Our method has been evaluated on the CAS-PEAL-R1, CMU PIE, and CUBIC FacePix databases. It has better performance than state-of-the-art methods for head pose estimation.

摘要：A new multi-species particle swarm optimization with a two-level hierarchical topology and the orthogonal learning strategy(OMSPSO) is proposed, which enhances the global search ability of particles and increases their convergence rates. The numerical results on 10 benchmark functions demonstrated the effectiveness of our proposed algorithm. Then, the proposed algorithm is presented to design a butterfly-shaped microstrip patch antenna. Combined with the HFSS solver, a butterfly-shaped patch antenna with a bandwidth of about 40.1% is designed by using the proposed OMSPSO. The return loss of the butterfly-shaped antenna is greater than 10 d B between 4.15 and 6.36 GHz. The antenna can serve simultaneously for the high-speed wireless computer networks(5.15–5.35 GHz) and the RFID systems(5.8 GHz).

摘要：Scavenging reactive oxygen species(ROS) by antioxidants is the important therapy to cerebral ischemia-reperfusion injury(CIRI) in stroke. The antioxidant with novel dual-antioxidant mechanism of directly scavenging ROS and indirectly through antioxidant pathway activation may be a promising CIRI therapeutic strategy. In our study, a series of chalcone analogues were designed and synthesized, and multiple potential chalcone analogues with dual antioxidant mechanisms were screened. Among these compounds, the most active 33 not only conferred cytoprotection of H2 O2-induced oxidative damage in PC12 cells through scavenging free radicals directly and activating NRF2/ARE antioxidant pathway at the same time, but also played an important role against ischemia/reperfusion-related brain injury in animals. More importantly, in comparison with mono-antioxidant mechanism compounds, 33 exhibited higher cytoprotective and neuroprotective potential in vitro and in vivo. Overall, our findings showed compound 33 couldemerge as a promising anti-ischemic stroke drug candidate and provided novel dual-antioxidant mechanism strategies and concepts for oxidative stress-related diseases treatment.

摘要：In order to break through the limitationof thelatitude/longitudegrid and hexagon grid, a new subdivision unit, Half-honeycomb Trapezoid, is proposed. Based on the summarization of the geometric properties and subdivision performance of Half-honeycomb Trapezoid, a new discrete global topographic grid system is established, and its compatibility with hexagonal grid is analyzed. At last, the visualization of multi-resolution global grid is achieved.

摘要：In the view of staff shortages and the huge inventory of products in the current market, we put forward a personnel scheduling model in the target of closing to the delivery date considering the parallelism. Then we designed a scheduling algorithm based on genetic algorithm and proposed a flexible parallel decoding method which take full use of the personal capacity. Case study results indicate that the flexible personnel scheduling considering the order-shop scheduling, machine automatic capabilities and personnel flexible in the target of closing to the delivery date optimize the allocation of human resources, then maximize the efficiency.

摘要：A new integer linear program （ILP） formulation was developed to solve the problem of shared cycles design for optical networks with and without wavelength conversion. Numerical studies show that the shared cycles design requires much lower redundancy compared with p-cycle design when the maximum number of spans allowed in the cycles is relatively small, but the difference in the redundancy between the two designs becomes smaller or disappears as the maximum number of spans allowed in the cycles increases.