Search

Hypertension is one of the most significant risk factors for atrial fibrillation (AF). However, few effective methods are available to support accurate prediction on the potential risk of atrial fibrillation among hypertensive patients currently. The aim of this paper is to illustrate a machine learning technology for constructing an atrial fibrillation intelligent prediction model. Eventually, the model can be employed to predict the risk of atrial fibrillation in hypertensive patients.A total of 2,067 diagnosed hypertensive patients (including 721 hypertensive patients complicated with atrial fibrillation) by Heart Center of Affiliated Zhongshan Hospital of Dalian University from January 2015 to January 2018 were enrolled in this study. As result, the atrial fibrillation prediction model was constructed based on the C5.0 decision tree classification algorithm. Moreover, compared with other machine learning classification algorithms, C5.0 has similar performance to random forest (RF), but is better than support vector machine(SVM), Logical Regression(LR), CHAID, and K nearest neighbor(KNN) classification algorithms. The proposed predict model has high accuracy of atrial fibrillation risk prediction for hypertension patients.

In this paper, diffusion bonding was adopted to join 9Cr martensitic/ferritic heat-resistant steels using an electrodeposited Ni interlayer with a thickness of 40 &mu, m. In addition, the effect of tempering treatment after diffusion bonding on the microstructure evolution and mechanical properties of the bonding joints was investigated. It was found that a transition region with face-centered cubic (FCC) structure was formed between the steel and Ni interlayer. The transition region was the solid solution of (&gamma, Fe,Ni) rich in Ni component, being related to martensite in the base metal by the Kurdjumov&ndash, Sachs (K-S) orientation relationship. No intermetallic compounds were detected at the bonding joints before and after tempering treatment. After tempering treatment, the transition region had higher dislocation density than other regions, due to the higher pinning effect of solute atoms acting on the dislocation than that of the matrix. Tensile tests indicated that tempering treatment improved the mechanical properties of the joint, since the samples after tempering treatment fractured in the base metal, whereas the specimens without tempering treatment fractured at the joint interface.

Three model powders defined as MP powders (milled pre-alloyed powders), mixed powders (MX, 50 wt.% MP powders + 50 wt.% Oxide-Dispersion Strengthened powders) and Oxide Dispersion Strengthened (ODS) powders (alloyed pre-alloyed powders with the addition of Ti and Y2O3) are obtained under identical ball milling parameters. These powders are then consolidated under same sintering condition by spark plasma sintering (SPS) in order to investigate the formation mechanism of martensite lath in the MP steel and the effect of Ti on the stability of ferrite. The results indicate that the addition of Y2O3 and Ti powders can act as friction material during the mechanical alloying process, thus promoting the refinement effect. The formation of martensite lath in the MP steel is attributed to the local Cr depletion resulted from the large amounts of M23C6 precipitation. Ti possesses a strong affinity to C and long range diffusion ability, which efficiently prevents the martensite lath formation and local Cr depletion. Present study supports the conclusion that the lack of martensite in the MX and ODS steel is due to the addition of Ti. Secondary phases in these steels are identified and analyzed as well.

In mobile cloud computing environments distributed k Nearest Neighbor (kNN) query is an important issue. We consider the problem of processing kNN query over large data sets where the index is jointly maintained by a set of machines in a computing cluster. The kNN query is a primitive operator that is widely used in many fields ranging from knowledge discovery, data mining and spatial databases etc. A scalable and distributed spatial data index plays an important role in conducting kNN query effectively. We can use different ways to conduct distributed indexes and kNN query processing by using MapReduce, i.e. R-tree and Grid-based index, etc. Nevertheless, R-tree is not compatible with parallelization, and Grid is a many-to-many index, which could potentially lead to content redundancy. In the paper, a distributed method of kNN queries applying MapReduce program model will be introduced. In the very beginning, I propose distributed methods which set up a novel distributed spatial data index: Inverted Voronoi Index that combines both inverted index and Voronoi diagram. Next, I propose a kNN queries processing algorithm, it is very efficient because it is based on Voronoi and uses MapReduce. Last but not least, I present the outcomes of extensive experiment that are gained by both real and simulated data sets which indicate efficiency and scalability of the proposed approach.