Hasil Pencarian  ::  Simpan CSV :: Kembali

Abstrak :
Data ekspresi gen adalah data yang menyajikan tingkat ekspresi gen pada kondisi yang berbeda-beda. Analisis data ekspresi gen microarray sangat penting karena memiliki aplikasi luas pada bidang biologi, medis, dan lain-lain. Dalam melakukan analisis data ekspresi gen, sebagian besar algoritma analisis membutuhkan data matriks yang lengkap. Sayangnya, beberapa data mungkin hilang karena kerusakan gambar, debu, dan kesalahan eksperimental. Oleh karena itu, metode imputasi missing value diperlukan untuk melakukan pemulihan pada data yang hilang tersebut. Penelitian ini mengembangkan suatu metode imputasi missing value, yaitu bicluster-based Bayesian principal component analysis dan robust least squares estimation dengan principal components (bi-BPCA-RLSP). Metode bi-BPCA-RLSP merupakan metode pengembangan dari bicluster-based robust least squares estimation dengan principal components (bi-RLSP). Pada metode bi-RLSP, tahap praimputasi untuk memperoleh matriks komplit sementara dilakukan dengan menggunakan metode row average. Namun, metode row average dinilai kurang baik dalam menggambarkan struktur keseluruhan data karena hanya menggunakan informasi dari baris yang mengandung missing value. Oleh karena itu, penelitian ini melakukan penggantian metode row average menjadi BPCA. BPCA menggunakan informasi korelasi dari seluruh data sehingga lebih baik dalam menggambarkan struktur keseluruhan data. Metode bi-BPCA-RLSP diterapkan pada data ekspresi gen garis sel kanker serviks dengan missing rate 1%, 5%, 10%, 15%, 20%, 25%, dan 30%. Performa metode bi-BPCA-RLSP diukur dengan menggunakan nilai normalized root mean squared error (NRMSE) dan dibandingkan dengan metode bi-RLSP. Hasil penelitian menunjukkan bahwa kinerja bi-BPCA-RLSP lebih baik daripada bi-RLSP karena nilai NRMSE pada bi-BPCA-RLSP lebih rendah dibandingkan bi-RLSP untuk setiap missing rate. ......Gene expression data is data that presents the level of gene expression under different conditions. Analysis of microarray gene expression data is very important because it has wide applications in the fields of biology, medicine, and others. In analyzing gene expression data, most of the analytical algorithms require a complete data matrix. Unfortunately, some data may be lost due to image corruption, dust, and experimental errors. Therefore, the missing value imputation method is needed to recover the lost data. This study developed a missing value imputation method, namely bicluster-based Bayesian principal component analysis and robust least squares estimation with principal components (bi-BPCA-RLSP). The bi-BPCA-RLSP method is a development method of bicluster-based robust least squares estimation with principal components (bi-RLSP). In the bi-RLSP method, the pre-imputation stage to obtain a temporary complete matrix is carried out using the row average method. However, the row average method is considered poor in describing the overall structure of the data because it only uses information from rows containing missing values. Therefore, this study replaced the row average method by BPCA. BPCA uses correlation information of all data so that it describes better the overall structure of the data. The bi-BPCA-RLSP method was applied to gene expression data of cervical cancer cell lines with missing rates of 1%, 5%, 10%, 15%, 20%, 25%, and 30%. The performance of the bi-BPCA-RLSP method was measured using the normalized root mean squared error (NRMSE) and compared with the bi-RLSP method. The results showed that bi-BPCA-RLSP performed better than bi-RLSP because the NRMSE value of bi-BPCA-RLSP was lower than bi-RLSP for each missing rate.

Abstrak :
This book presents the second volume of Piola’s original Italian text together with the English-language translation and comments, showing convincingly that Gabrio Piola’s work must still be regarded as a modern theory. Gabrio Piola’s work has had an enormous impact on the development of applied mathematics and continuum mechanics. As such, a committee of scientific experts took it upon themselves to translate his complete works. In a second step, they commented on Piola’s work and compared it to modern theories in mechanics in order to stress Piola’s impact on modern science and prove and confirm that he achieved significant milestones in applied mathematics.

Abstrak :
Papers appearing in this volume are the Invited Talks given at ICIAM 2003, the 5th International Congress of Industrial and Applied Mathematics, held in Sydney over the period July 7 to 11, 2003. The Congress celebrates and describes the contributions of applied mathematics -- as an intellectual creation in its own right, as a foundation stone of technological development, and as an indispensable collaborative partner for other scientific disciplines. These aspects of applied mathematics have held true since the dawn of civilization. They will remain just as important in the new millennium. The authors of these papers are all acknowledged masters of their fields, having been chosen through a rigorous selection process by a distinguished International Program Committee. This volume presents an overview of contemporary applications of mathematics, with the coverage ranging from the rhythms of the nervous system, to optimal transportation, elasto-plasticity, computational drug design, hydrodynamic and meteorological modeling, and valuation in financial markets.

Abstrak :
Metoda Elemen hingga adalah salah satu metoda numerik yang dipakai untuk menyelesaikan persamaan differensial. Pada tugas akhir ini akan dibahas penggunaan Metoda Elemen Hingga dalam menyelesaikan masalah aliran fluida viskous inkompresibel dimana persamaan atur dari fluida ini adalah persamaan yang multivariabel.

Abstrak :
This textbook presents basic numerical methods and applies them to a large variety of physical models in multiple computer experiments. Classical algorithms and more recent methods are explained. Partial differential equations are treated generally comparing important methods, and equations of motion are solved by a large number of simple as well as more sophisticated methods. Several modern algorithms for quantum wavepacket motion are compared. The first part of the book discusses the basic numerical methods, while the second part simulates classical and quantum systems. Simple but non-trivial examples from a broad range of physical topics offer readers insights into the numerical treatment but also the simulated problems. Rotational motion is studied in detail, as are simple quantum systems. A two-level system in an external field demonstrates elementary principles from quantum optics and simulation of a quantum bit. Principles of molecular dynamics are shown. Modern bounda ry element methods are presented in addition to standard methods, and waves and diffusion processes are simulated comparing the stability and efficiency of different methods. A large number of computer experiments is provided, which can be tried out even by readers with no programming skills. Exercises in the applets complete the pedagogical treatment in the book. In the third edition Monte Carlo methods and random number generation have been updated taking recent developments into account. Krylov-space methods for eigenvalue problems are discussed in much more detail. The wavelet transformation method has been included as well as simple applications to continuum mechanics and convection-diffusion problems. Lastly, elementary quantum many-body problems demonstrate the application of variational and Monte-Carlo methods.

Abstrak :
Equity Risk Premium (ERP) adalah besarnya kompensasi yang didapat oleh investor yang berani berinvestasi pada risiko yang lebih tinggi. Setiap portfolio investasi membawa tingkat risiko yang berbeda-beda dan ERP yang menggambarkan besaran dari harga atas risiko tersebut. Radulescu-Pele (2014) menawarkan ide bahwa runtun waktu makroekonomi dapat digunakan untuk menaksir ERP. Namun, runtun makoekonomi tidak dapat langsung digunakan dalam memodelkan ERP dalam suatu model regresi. Runtun makroekonomi mengandung growth component dan cyclical component. Sehingga, terlebih dahulu harus dilakukan dekomposisi untuk memisahkan kedua komponen tersebut. Salah satu metode dekomposisi yang dapat digunakan adalah Hodrick-Prescott Filter. Dalam metode HP Filter, pemilihan nilai parameter pemulusan yang tepat menjadi salah satu kendala. Pemilihan besarnya parameter pemulusan dapat dilakukan dengan memperhatikan interval waktu pengamatan dari runtun waktu makroekonomi yang digunakan dalam memodelkan ERP.

---

Equity Risk Premium (ERP) is the increasing of the risk for the expected additional return. Every portfolio brings different rate of risk and ERP presents the compensatition of the risk.. Radulescu and Pele (2014) suggested that macroeconomics series can be used for estimating ERP. But the series of macroeconomics can not be used directly. The variables have to be decomposed first to separate growth component and cyclical component. One of the decomposition?s method is Hodrick-Prescott Filter. In this method, the selection of the smoothness parameter could be a problem. The selection of smoothness parameter can be based on the length of macroeconomics series? observation in modeling the ERP.