Support vector machine (SVM) is a popular supervised learning algorithm based on margin maximization. It has a high training cost and does not scale well to a large number of data points. We propose a multiresolution algorithm MRH-SVM that trains SVM on a hierarchical data aggregation structure, which also serves as a common data input to other learning algorithms. The proposed algorithm learns SVM models using high-level data aggregates and only visits data aggregates at more detailed levels where support vectors reside. In addition to performance improvements, the algorithm has advantages such as the ability to handle data streams and datasets with imbalanced classes. Experimental results show significant performance improvements in comparison with existing SVM algorithms.
In this work, the switching nonlinear dynamics of a Fabry-Perot etalon are studied. The method used to complete the solution of the differential equations for the nonlinear medium. The Debye relaxation equations solved numerically to predict the behavior of the cavity for modulated input power. The response of the cavity filled with materials of different response time is depicted. For a material with a response time equal to = 50 ns, the cavity switches after about (100 ns). Notice that there is always a finite time delay before the cavity switches. The switch up time is much longer than the cavity build-up time of the corresponding linear cavity which was found to be of the order of a few round-trip ti
... Show MoreThis paper present a simple and sensitive method for the determination of DL-Histidine using FIA-Chemiluminometric measurement resulted from oxidation of luminol molecule by hydrogen peroxide in alkaline medium in the presence of DL-Histidine. Using 70?l. sample linear plot with a coefficient of determination 95.79% for (5-60) mmol.L-1 while for a quadratic relation C.O.D = 96.44% for (5-80) mmol.L-1 and found that guadratic plot in more representative. Limit of detection was 31.93 ?g DL-Histidine (S/N = 3), repeatability of measurement was less that 5% (n=6). Positive and negative ion interferances was removed by using minicolume containing ion exchange resin located after injection valve position.
Background: Atrioventricular nodal reentrant tachycardia (AVNRT) is the commonest regular supraventricular tachyarrhythmia. Ablation in the area of slow pathway (SP) has been successfully implemented in every day clinical electrophysiological practice for more than 20 years. Although the procedure is generally regarded as effective and safe, data on long-term effects and predictors of success or failure are incomplete.
This studies p- CuO / n - Si hete-rojunction was deposited by high vacuum thermal evaporation of Copper subjected to thermal oxidation at 300 oC on silicon. Surface morphology properties of The optical properties concerning the transmission spectra were studies for prepared thin films. this structure have been studied. XRD anaylsis discover that the peak at (𝟏𝟏𝟏-) and (111) plane are take over for the crystal quality of the CuO films. The band gap of CuO films is found to be 1.54 eV. The average grain size of is measured from AFM analysis is around 14.70 nm. The responsivity photodetector after deposited CuO appear increasing in response
Finite Element Approach is employed in this research work to solve the governing differential equations related to seepage via its foundation's dam structure. The primary focus for this reason is the discretization of domain into finite elements through the placement of imaginary nodal points and the discretization of governing equations into an equation system; An equation for each nodal point or part, and unknown variables are solved. The SEEP / W software (program) is a sub-program of the Geo-Studio software, which is used by porous soil media to compensate for the problems of seepage. To achieve the research goals, a study was carried out on Hemrin dam, which located in the Diyala River 100 km northeast of Baghdad, Iraq. Thus, o
... Show MoreThe goal of this work is demonstrating, through the gradient observation of a of type linear ( -systems), the possibility for reducing the effect of any disturbances (pollution, radiation, infection, etc.) asymptotically, by a suitable choice of related actuators of these systems. Thus, a class of ( -system) was developed based on finite time ( -system). Furthermore, definitions and some properties of this concept -system and asymptotically gradient controllable system ( -controllable) were stated and studied. More precisely, asymptotically gradient efficient actuators ensuring the weak asymptotically gradient compensation system ( -system) of known or unknown disturbances are examined. Consequently, under convenient hypo
... Show More