GSPL: A Succinct Kernel Model for Group-Sparse Projections Learning of Multiview Data

This paper explores a succinct kernel model for Group-Sparse Projections Learning (GSPL), to handle multiview feature selection task completely. Compared to previous works, our model has the following useful properties: 1) Strictness: GSPL innovatively learns group-sparse projections strictly on multiview data via ‘2;0-norm constraint, which is different with previous works that encourage group-sparse projections softly. 2) Adaptivity: In GSPL model, when the total number of selected features is given, the numbers of selected features of different views can be determined adaptively, which avoids artificial settings. Besides, GSPL can capture the differences among multiple views adaptively, which handles the inconsistent problem among different views. 3) Succinctness: Except for the intrinsic parameters of projection-based feature selection task, GSPL does not bring extra parameters, which guarantees the applicability in practice. To solve the optimization problem involved in GSPL, a novel iterative algorithm is proposed with rigorously theoretical guarantees. Experimental results demonstrate the superb performance of GSPL on synthetic and real datasets.

Comparison of Prediction Models for Mortality Related to Injuries from Road Traffic Accidents after Correcting for Undersampling

In this study, four models—logistic regression (LR), random forest (RF), linear support vector machine (SVM), and radial basis function (RBF)-SVM—were compared for their accuracy in determining mortality caused by road traffic injuries. They were tested using five years of national-level data from the Korea Disease Control and Prevention Agency’s (KDCA) National Hospital Discharge In-Depth Survey (2013 through to 2017). Model performance was measured for accuracy, precision, recall, F1 score, and Brier score metrics using classification analysis that included characteristics of patients, accidents, injuries, and illnesses. Due to the number of variables and differing units, the rates of survival and mortality related to road traffic accidents were imbalanced, so the data was corrected and standardized before the classification models’ performances were compared. Using the importance analysis, the main diagnosis, the type of injury, the site of the injury, the type of injury, the operation status, the type of accident, the role at the time of the accident, and the sex were selected as the analysis factors. The biggest contributing factor was the role in the accident, which is the driver, and the major sites of the injuries were head injuries and deep injuries. Using selected factors, comparisons of the classification performance of each model indicated RBF-SVM and RF models were superior to the others. Of the SVM models, the RBF kernel model was superior to the linear kernel model; it can be inferred that the performance of the high-dimensional transformed RBF model is superior when the dimension is complex because of the use of multiple variables. The findings suggest there are limitations to analyses involving imbalanced, multidimensional original data, such as data on road traffic mortality. Thus, analyses must be performed after imbalances are corrected.

NEW SOLUTION TO THE DROPLET KERNEL MODEL

The droplet model of the nucleus is revived, for which an exact solution for an incompressible fluid is obtained using the hydrodynamic potential solution obtained from the Schrödinger equation. Moreover, for an incompressible fluid, there are formulas for the pressure or potential. There is the main part of the hydrodynamic potential, which is obtained by replacing the modulus of the inverse difference of vectors by the difference in moduli of the values of the vectors. The bulk of the potential is expressed in a finite formula with singularities. A formula is obtained for the integral containing the modulus of the difference between the exact values of the vectors minus the main part of the potential. This difference defines a continuous correction with the features taken into account. The main part of the potential at the boundary of the nucleus turned out to be infinitely large with an imaginary part, locking particles in the nucleus. In this case, the real part of the main potential decreases with decreasing radius, becomes negative, and determines the bound state. At half the radius of the nucleus, there is a linear term along the radius. At the zero radius, there is an infinite negative potential with an imaginary part. An expression for the quantum of the emitted energy is obtained. Note that the added mass was not used due to the rotational regime of the nucleus. An algorithm for calculating the spectrum of the kernel is proposed, and each state of the action of the kernel sn corresponds to n calculated frequencies, determined by n angles in the configuration space. The main space is n + 1 dimensional, and each dimension of space has its own energy. But without special means, the potential of the nucleus tends to infinity. It is necessary to introduce the imaginary degree of roughness of the corners, in expressions containing singularities, then the infinities disappear.

Relative occurrence rates of terrestrial planets orbiting FGK stars

This paper aims to derive a map of relative planet occurrence rates that can provide constraints on the overall distribution of terrestrial planets around FGK stars. Based on the planet candidates in the Kepler DR25 data release, I first generate a continuous density map of planet distribution using a Gaussian kernel model and correct the geometric factor that the discovery space of a transit event decreases along with the increase of planetary orbital distance. Then I fit two exponential decay functions of detection efficiency along with the increase of planetary orbital distance and the decrease of planetary radius. Finally, the density map of planet distribution is compensated for the fitted exponential decay functions of detection efficiency to obtain a relative occurrence rate distribution of terrestrial planets. The result shows two regions with planet abundance: one corresponds to planets with radii between 0.5 and 1.5 R⊕ within 0.2 AU, the other corresponds to planets with radii between 1.5 and 3 R⊕ beyond 0.5 AU. It also confirms the features that may be caused by atmospheric evaporation: there is a vacancy of planets of sizes between 2.0 and 4.0 R⊕ inside of ∼ 0.5 AU, and a valley with relatively low occurrence rates between 0.2 and 0.5 AU for planets with radii between 1.5 and 3.0 R⊕.

Biresponses Kernel Nonparametric Regression: Inflation and Economic Growth

The relation between inflation and economic growth is interesting to observe. To maintain the inflation rate, two factors should be taken into account, namely keeping the economic pulse at its optimal rate and keeping people's purchasing power from decreasing. Many factors influence the inflation and economic growth of a nation; one of which is the national bank interest rate. Since the data of inflation are closely related to economic growth, this study aims at modelling the data of inflation rate and economic growth of Central Java Province in Indonesia using bi-response kernel regression. Employing the data from the first trimester of 2007 up to those from the second trimester of 2019 which were processed using kernel Gauss, the best model to minimise the value of GCV was obtained with optimum h for inflation model amounting to 0.12 and 81.75 for economic growth model. The model performance was excellent because the MAPE out sample was less than 10%. The biresponses kernel model is better than the linear biresponses model in terms of GCV, MSE, R2, and MAPE values.

Klasifikasi Mahasiswa HER Berbasis Algoritma SVM dan Decision Tree

<p>Mahasiswa di setiap perguruan tinggi dituntut untuk memperoleh pengetahuan dan keterampilan yang memenuhi syarat dengan prestasi akademik. Hasil dari pembelajaran mahasiswa didapat dari ujian teori dan praktek, setiap mahasiswa wajib menuntaskan nilai sesuai kriteria kelulusan minimum dari masing-masing dosen pengajar, jika dibawah batas minimum maka mahasiswa mengikuti her. Her adalah salah satu cara untuk menuntaskan kriteria kelulusan minimum. Mahasiswa yang mengikuti her setiap semesternya hampir mencapai angka yang relatif tinggi dari jumlah seluruh mahasiswa. Untuk mengurangi jumlah mahasiswa yang mengikuti her maka dibutuhkan sebuah metode yang dapat mengurangi hal tersebut, dengan metode <em>Support Ve</em><em>c</em><em>tor Machine</em> (SVM) dan <em>Decision Tree </em>(DT). SVM dan DT adalah salah satu metode klasifikasi <em>supervised learning</em>. Oleh karena itu, dalam penelitian ini menggunakan SVM dan DT. SVM dapat menghilangkan hambatan pada data, memprediksi, mengklasifikasikan dengan sampling kecil dan dapat meningkatkan akurasi dan mengurangi kesalahan. Klasifikasi data siswa yang melakukan her/peningkatan dengan mengimprovisasi model kernel untuk visualisasi termasuk bar, histogram, dan sebaran<em> </em>begitu juga<em> Decision Tree </em>mempunyai kelebihan tersendiri. Dari hasil penelitian ini telah didapatkan akruasi dan presisi model DT lebih besar dibandingkan dengan SVM, akan tetapi untuk <em>recall </em>DT lebih kecil dibandingkan SVM.</p><p> </p><p><em><strong>Abstract</strong></em></p><p><em><strong><br /></strong></em></p><p class="Abstract"><em>Students in each tertiary institution are required to obtain knowledge and skills that meet the requirements with academic achievement. The results of student learning are obtained from the theory and practice exams, each student is required to complete grades according to the minimum graduation criteria of each teaching lecturer, if below the minimum limit then students take remedial. Remedial is one way to complete the minimum passing criteria. Students who take remedial every semester almost reach a relatively high number of the total number of students. To reduce the number of students who take remedial, a method that can reduce this is needed, with the Support Vector Machine (SVM) and Decision Tree (DT) methods. SVM and DT are one of the supervised learning classification methods. Therefore, in this study using SVM and DT. SVM can eliminate barriers to data, predict, classify with small sampling and can improve accuracy and reduce errors. Data classification of students who do remedial/improvements by improving the kernel model for visualization including bars, histograms, and distributions as well as the Decision Tree has its own advantages. From the results of this study it has been obtained that the accuracy and precision of DT models is greater than that of SVM, but for recall DT is smaller than SVM.</em></p><p><em><strong><br /></strong></em></p>