Claim Missing Document
Check
Articles

Found 2 Documents
Search
Journal : Jurnal Matematika

Perbandingan ELM dan Double Exponential Smoothing Untuk Meramalkan PDRB Di Provinsi NTT Laura Liokelly Toron; Yudi Setyawan; Noviana Pratiwi
Jurnal Matematika Vol 12 No 1 (2022)
Publisher : Publisher : Mathematics Department, Faculty of Mathematics and Natural Sciences, Udayana University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24843/JMAT.2022.v12.i01.p147

Abstract

Abstract: Gross Regional Domestic Product is the total number of goods and services produced by production units of all economic sectors of a particular region during one year. BPS NTT noted that the economic growth rate of NTT in 2020 experienced a contraction of -0.83% from 5.24% in the previous year, so this study aims to predict NTT's GRDP using the ELM method and Holt's Double Exponential Smoothing. ELM is an artificial neural network that has one hidden layer that is applied through training and testing process, then involves a binary sigmoid activation function and a Moore Penrose Pseudo Inverse matrix to get the output weight used to predict. DES Holt is a forecasting method that pays attention to trend data plots and uses two parameters in its calculations. The results of the forecasting research show that the ELM method with a proportion of 80%:20% is the best method for predicting the GRDP of NTT. The ELM method produces quarterly GRDP values in 2021, which are 17493.19754, 18154.80753, 18712.02153, and 18822.97416 (billion rupiah) with 4 input neurons, 12 hidden layer neurons, 1 output neuron and the MAPE value is 0.7968% which is smaller than DES Holt.
Analisis Positioning Merk Laptop dengan Menggunakan Metode MDS Nonmetrik dan CA Maria Romaana Ona Sain; Yudi Setyawan; Rokhana Dwi Bekti
Jurnal Matematika Vol 12 No 2 (2022)
Publisher : Publisher : Mathematics Department, Faculty of Mathematics and Natural Sciences, Udayana University

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24843/JMAT.2022.v12.i02.p152

Abstract

Abstract: This study aims to determine the positioning of laptop brands on attributes based on the perceptions and preferences of IST AKPRIND students. The method used is nonmetric multidimensional scaling and correspondence analysis. The results showed that: multidimensional scaling of perception data, quadrant I was occupied by HP and Dell, quadrant II was occupied by Toshiba, quadrant III was occupied by Acer, Lenovo, Asus, and quadrant IV was occupied by Apple. Then multidimensional scaling preference data, it is known that quadrant I is occupied by storage and price attributes, quadrant II is occupied by Apple with attributes of laptop resistance to damage, feature set, RAM, processor, quadrant III is occupied by brand image and warranty attributes, and quadrant IV is occupied by HP, Dell, Toshiba, Acer, Lenovo, Asus, and there are no attributes in quadrant IV. Using correspondence analysis, it is known that quadrant I is occupied by Apple with price attributes, quadrant II is occupied by Toshiba with attributes of brand image, processor, RAM, feature set, quadrant III is occupied by HP, Dell, Lenovo, Acer, Asus with attributes of laptop resistance to damage, quadrant IV is occupied storage and warranty attributes. There is no laptop in quadrant IV.