Unrestricted Random Sampling of data batch to improve the efficiency of neural networks

The aim of this study was to reveal cranio-spinal differences between skeletal classification using convolutional neural networks (CNNs). Transverse and longitudinal cephalometric images of 832 patients were used for training and testing of CNNs (365 males and 467 females). Labeling was performed such that the jawbone was sufficiently masked, while the parts other than the jawbone were minimally masked. DenseNet was used as the feature extractor. Five random sampling crossvalidations were performed for two datasets. The average and maximum accuracy of the five crossvalidations were 90.43% and 92.54% for test 1 (evaluation of the entire posterior–anterior (PA) and lateral cephalometric images) and 88.17% and 88.70% for test 2 (evaluation of the PA and lateral cephalometric images obscuring the mandible). In this study, we found that even when jawbones of class I (normal mandible), class II (retrognathism), and class III (prognathism) are masked, their identification is possible through deep learning applied only in the cranio-spinal area. This suggests that cranio-spinal differences between each class exist.

Download Full-text

Analisis Pengaruh Biaya Terhadap Loyalitas Pengguna Mobile Banking Jenius Menggunakan Model Neural Networks

Applied Information System and Management (AISM) ◽

10.15408/aism.v2i2.20163 ◽

2021 ◽

Vol 2 (2) ◽

Author(s):

Ginanjar Ramadhan ◽

Rinda Hesti Kusumaningtyas ◽

Aries Susanto Hasan Taufik

Keyword(s):

Neural Networks ◽

Customer Satisfaction ◽

Customer Loyalty ◽

Random Sampling ◽

Perceived Value ◽

Perceived Quality ◽

Mobile Banking

Perbedaan biaya layanan seperti biaya transfer dan biaya administrasi yang diterapkan oleh beberapa institusi menimbulkan beberapa sikap dari nasabah, salah satunya terkait loyalitas pelanggan. Loyalitas merupakan suatu perilaku yang ditunjukkan dengan penggunaan berulang-ulang yang didasarkan pada pengambilan keputusan. Mobile banking Jenius merupakan layanan perbankan yang diberikan oleh Bank Tabungan Pensiunan Nasional (BTPN). Belum diketahui faktor-faktor yang mempengaruhi loyalitas pengguna, membuat mobile banking Jenius belum maksimal dalam pengembangan produk, biaya juga menjadi salah satu faktor yang belum diketahui pengaruhnya sehingga keputusan perubahan biaya tidak diketahui. Status loyalitas pengguna mobile banking Jenius juga belum diketahui, sehingga keputusan untuk membuat program loyalitas belum dilakukan secara maksimal. Penelitian ini bertujuan untuk mengetahui faktor-faktor yang mempengaruhi loyalitas pengguna mobile banking Jenius menggunakan model Neural Networks yang dimodifikasi dengan menambahkan variabel cost. Populasi penelitian ini adalah pengguna mobile banking Jenius dengan teknik pengambilan sampel yaitu random sampling. Sementara analisis data menggunakan pendekatan PLS-SEM dengan bantuan aplikasi Smart PLS 3.0. Hasil dari penelitian ini adalah switching barries, commitment, perceived quality, empathy, cost mempengaruhi perceived value dan customer satisfaction. Perceived value dan customer satisfaction berpengaruh terhadap trust dan customer loyalty. Hasil dari penelitian ini diharapkan dapat menjadi salah satu bahan pertimbangan pengambilan keputusan Bank BTPN untuk meningkatkan loyalitas pengguna mobile banking Jenius.

Download Full-text

Selective Sampling and Orientation of Non-Homogeneous Tissues

Proceedings, annual meeting, Electron Microscopy Society of America ◽

10.1017/s0424820100100238 ◽

1968 ◽

Vol 26 ◽

pp. 98-99

Author(s):

C. C. Clawson ◽

L. W. Anderson ◽

R. A. Good

Keyword(s):

Electron Microscopy ◽

Electron Microscope ◽

Light Microscopy ◽

Random Sampling ◽

New Method ◽

Microscope Examination ◽

Small Areas ◽

Selective Sampling ◽

Large Numbers ◽

Specific Orientation

Investigations which require electron microscope examination of a few specific areas of non-homogeneous tissues make random sampling of small blocks an inefficient and unrewarding procedure. Therefore, several investigators have devised methods which allow obtaining sample blocks for electron microscopy from region of tissue previously identified by light microscopy of present here techniques which make possible: 1) sampling tissue for electron microscopy from selected areas previously identified by light microscopy of relatively large pieces of tissue; 2) dehydration and embedding large numbers of individually identified blocks while keeping each one separate; 3) a new method of maintaining specific orientation of blocks during embedding; 4) special light microscopic staining or fluorescent procedures and electron microscopy on immediately adjacent small areas of tissue.

Download Full-text