Many mixed datasets with both numerical and categorical attributes have been collected in various fields, including medicine, biology, etc. Designing appropriate similarity measurements plays an important role in clustering these datasets. Many traditional measurements treat various attributes equally when measuring the similarity. However, different attributes may contribute differently as the amount of information they contained could vary a lot. In this paper, we propose a similarity measurement with entropy-based weighting for clustering mixed datasets. The numerical data are first transformed into categorical data by an automatic categorization technique. Then, an entropy-based weighting strategy is applied to denote the different importances of various attributes. We incorporate the proposed measurement into an iterative clustering algorithm, and extensive experiments show that this algorithm outperforms OCIL and K-Prototype methods with 2.13% and 4.28% improvements, respectively, in terms of accuracy on six mixed datasets from UCI.

The Ambient Assisted Living (AAL) systems use sensors to detect the daily behavior of older adults and provide necessary assistance based on changes in their cognitive status and physical functions, thus enabling older adults to maintain their independence at home. However, the effectiveness of the AAL systems depends on the accuracy of the data provided by sensors. Namely, when a human error or a hardware failure occurs, the activity recognition model can become inaccurate. This inaccuracy hinders the identification of critical and potentially life-threatening activities. Although there are many methods for cleaning sensor data, there is no method for binary sensors deployed in smart homes. By considering noisy sensor events and unintentional forgetting of turning off the device, this paper proposes two clustering-based methods for denoising and splitting binary sensor events to address possible inaccuracy due to the two mentioned problems. The effectiveness of the proposed methods is verified by the experiments using four machine learning models and three real-world smart home datasets and adopting different sensor configurations. The experimental results demonstrate that compared to the original unprocessed datasets, by combining the two proposed methods, the average accuracy and F-measure are improved by 15.00% and 17.25%, respectively.

目前,高校数据分析类课程教学模式大都采用纯"知识点"教学法,往往导致教学内容比较呆板.为改进教学内容呆板的缺点,一些高校在教学中引入案例教学法,但带来的结果是教学内容连贯性不足.针对以上问题,面向健康计算领域应用编排教学内容,对传统案例教学方法进行改进,采用真实世界数据主线贯穿知识点的案例教学法,使得教学内容活泼连贯.最后,针对实验模式建设现状提出相应解决方案.

采用格子波尔兹曼方法结合有限差分方法求解黏弹性流体的运动方程,非线性有限元方法求解柔性丝线的运动方程,浸没边界法处理流固耦合,数值求解了柔性丝线在均匀来流中因流固耦合作用产生的自由拍动问题.通过与牛顿流情况的对比,重点考察了流体弹性效应对丝线拍动特性的影响.计算结果发现,对于流体弹性效应较弱的情况(We<20),柔性丝线从稳定(静止)模态过渡为周期性拍动模态的临界质量比随着We数的增加显著增大;而对于弹性效应较强的情况(We>20),临界质量比随着We数的增加渐近地趋于一常值.另外还发现,当质量比给定时,拍动丝线的阻力系数、拍动幅值和拍动频率均随We的增加而减小.以上发现表明,流体弹性效应的增强对丝线拍动和尾迹流动失稳脱涡具有明显的抑制作用.

Microarray data suffer from missing values for various reasons, including insufficient resolution, image noise, and experimental errors. Because missing values can hinder downstream analysis steps that require complete data as input, it is crucial to be able to estimate the missing values. In this study, we propose a Global Learning with Local Preservation method (GL2P) for imputation of missing values in microarray data. GL2P consists of two components: a local similarity measurement module and a global weighted imputation module. The former uses a local structure preservation scheme to exploit as much information as possible from the observable data, and the latter is responsible for estimating the missing values of a target gene by considering all of its neighbors rather than a subset of them. Furthermore, GL2P imputes the missing values in ascending order according to the rate of missing data for each target gene to fully utilize previously estimated values. To validate the proposed method, we conducted extensive experiments on six benchmarked microarray datasets. We compared GL2P with eight state-of-the-art imputation methods in terms of four performance metrics. The experimental results indicate that GL2P outperforms its competitors in terms of imputation accuracy and better preserves the structure of differentially expressed genes. In addition, GL2P is less sensitive to the number of neighbors than other local learning-based imputation methods.

Direct numerical simulations (DNS ) using the spectral method to solve the three-dimensional incompressible Navier-Stokes equations in cylindrical coordinates were performed to study the influence of radius ratio on the turbulent Taylor-Couette (TC) flows between two rotating concentric cylinders .As the TC flows are well accepted as imposition of turbulence fluctuations on the organized large-scale Taylor vortices (TV ) ,averaging along the azimuthal direction and in time was used to separate the large-scale TVs from the TC flow and thus divide the turbulent motions into the TVs-induced and the randomly fluctuating parts . The contributions to the velocity fluctuation intensities and turbulent kinetic energy , coming from the TVs-induced and turbulent fluctuating motions , were examined for different radius ratios .Moreover ,the budget equations of Reynolds stresses were calculated to elucidate the influence of radius ratio on the dynamical features of turbulence energy production ,dissipation and redistribution .It was demonstrated that for a narrow-gap TC flow with a low radius ratio ,the relatively stronger random fluctuating motions have great contributions to the statistics ,as compared to the TVs-induced fluctuations which ,however , becomes dominant in a wide-gap TC flow with a high radius ratio . Additionally , increasing radius ratios leads to an increase in the shear rates of mean flow in the near outer wall region . As a consequence ,the TC flow with a high radius ratio behaves in a similar fashion to the planar Couette flow .

Turbulent channel flows with consideration of the buoyancy effect of the bubble phase is investigated by means of the Direct Numerical Simulation (DNS). This two-phase system is solved by a two-way coupling Lagrangian-Eulerian approach. The Reynolds number based on the friction velocity and the half-width of the channel is 194, and the gravitational acceleration varies from −0.5 to 0.5, ranging from the upflow to the downflow cases. This study aims to reveal the influence of buoyancy on the turbulence behavior and the bubble motion. Some typical statistical quantities, including the averaged velocities and velocity fluctuations for the fluid and bubble phases, as well as the flow structures of the turbulence fluctuations, are analyzed.

Direct numerical simulations have been used to study turbulent bubbly flows in a vertical channel.The effect of bubble motions on the turbulent fluctuations was examined for the turbulent bubbly flows when the direction of gravity is aligned with/against the mean flow.Turbulence statistics for the turbulent bubbly channel flows were given and compared with those of pure-shear channel flow.Some typical results,such as the concentration and velocity fluctuations of bubble phase,and the near-wall coherent structures,were exhibited and analyzed.It is found that when the direction of gravity is against the mean flow,the turbulence statistics are enhanced by the bubble motion.The effect of bubble motion is to suppress the turbulent fluctuation when the direction of gravity is aligned with the mean flow.