Suliadi
Statistika, Fakultas Matematika dan Ilmu Pengetahuan Alam

Published : 3 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 3 Documents
Search

Diagram Kendali X Exponentially Weighted Moving Average yang Meminimalkan Median Run Length pada Data Panjang Pewarna Plastik Shania Wilda Fitris; Suliadi
Bandung Conference Series: Statistics Vol. 3 No. 1 (2023): Bandung Conference Series: Statistics
Publisher : UNISBA Press

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29313/bcss.v3i1.5513

Abstract

Abstract. The main purpose of statistical quality control is to quickly investigate whether special causes or process shifts has occurred so that an investigation of the process and corrective action can be taken. One of the tools for statistical quality control is the control chart. The tool that can be used to measure the performance of the control chart is the average run length (ARL). The disadvantage is that the run length distribution is skewed when the process is in control or slightly out of control, thus ARL’s interpretation of the performance of the control chart is less meaningful. You et al. (2016) introduced EWMA control chart based on the median run length (MRL), that is the Xbar EWMA control chart that minimizes the median run length (MRL) and expected median run length (EMRL). This method is more informative and reliable and is not affected by the slope of the run length distribution compared to ARL. In this study we applied the method to the the plastic colorant length data of Company X. From the research conducted, it was concluded that plastic colorant length data of Company X was statistically in control both in phase-I and phase-II data. Abstrak. Tujuan utama dari pengendalian kualitas statistik yaitu menyelidiki dengan cepat apakah terjadi penyebab khusus atau pergeseran proses sedemikian sehingga penyelidikan terhadap proses tersebut dan tindakan perbaikan dapat dilakukan. Salah satu alat untuk pengendalian kualitas statistik yaitu diagram kendali. Alat yang bisa digunakan untuk melihat kinerja diagram kendali adalah average run length (ARL). Kelemahannya adalah distribusi run length miring ketika proses in control atau sedikit out of control, sehingga interpretasi ARL mengenai kinerja dari diagram kendali kurang berarti. Maka You dkk. (2016) memperkenalkan alternatif dari penggunaan diagram kendali EWMA berdasarkan median run length (MRL), yaitu mengunakan diagram kendali Xbar EWMA yang meminimalkan median run length (MRL) dan expected median run length (EMRL), karena metode ini lebih informatif dan reliabel serta tidak dipengaruhi oleh kemiringan distribusi run length dibandingkan menggunakan ARL. Dalam penelitian ini kami menerapkan metode tersebut terhadap data panjang pewarna plastik Perusahaan X. Dari penelitian yang dilakukan disimpulkan bahwa data panjang pewarna plastik Perusahaan X terkendali secara statistik baik pada data fase-I maupun data fase-II.
Rekomendasi Destinasi Wisata di Indonesia Menggunakan Metode Item2Vec Aisha Kusuma Putri; Suliadi
Bandung Conference Series: Statistics Vol. 3 No. 1 (2023): Bandung Conference Series: Statistics
Publisher : UNISBA Press

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29313/bcss.v3i1.5666

Abstract

Abstract. Everyone choosing an item can be different, thus recommendations will be difficult to be given when many peoples have their own preferences. Hence, we need a method that can provide recommendations to many peoples, one of them is Item2Vec. Item2Vec is an item recommendation method that only uses user rating data in its algorithm. The items recommended to users are items that are similar to items previously liked. In its algorithm, Item2Vec is built with collaborative filtering (CF) and Word2Vec Skip Gram with Negative Sampling (SGNS). Through Word2Vec SGNS, it represents items into a vector of numbers through a neural network process, that is feedforward, that minimizing the objective function and backpropagation. Through CF, the vector for each item will be calculated using cosine similarity as the similarity value between items. The highest cosine similarity value results are interpreted as the most recommended items for each item and otherwise. Item2Vec has greater accuracy and more effective performance than recommendation systems built by Singular Value Decomposition (SVD) with CF. The objective of this research is to provide recommendation for tourist destination in Indonesia, by using Item2Vec method. This method makes it easier for tourists to choose tourist destinations and also makes it easier for travel agents to offer tourist destinations for promotion purposes. The Item2Vec method produces recommendations for tourist destinations in Indonesia with an accuracy of 61% and is considered to be performing quite well. Abstrak. Keinginan setiap orang dalam memilih suatu item dapat berbeda-beda, sehingga rekomendasi akan sulit diberikan ketika banyak orang memiliki keinginannya masing-masing. Maka, diperlukannya metode yang dapat memberikan rekomendasi kepada banyak orang salah satunya yaitu Item2Vec. Item2Vec merupakan sistem rekomendasi item yang hanya menggunakan data rating pengguna dalam algoritmanya. Item yang direkomendasikan kepada pengguna yaitu item yang memiliki kemiripan dengan item yang disukai sebelumnya. Dalam algoritmanya, Item2Vec dibangun dengan collaborative filtering (CF) dan Word2Vec Skip Gram with Negative Sampling (SGNS). Melalui Word2Vec SGNS, direpresentasikannya item ke dalam vektor angka melalui proses neural network yaitu feedforward, meminimumkan fungsi objektif dan backpropagation. Melalui CF, vektor setiap item dihitung menggunakan cosine similarity sebagai nilai kemiripan antar item. Hasil nilai cosine similarity tertinggi diartikan sebagai item yang paling direkomendasikan untuk setiap item dan juga sebaliknya. Item2Vec memiliki akurasi lebih besar dan kinerja lebih efektif dibandingkan sistem rekomendasi yang dibangun oleh Singular Value Decompotion (SVD) dengan CF. Tujuan dari penelitian ini adalah memberikan rekomendasi destinasi wisata di Indonesia, dengan menggunakan metode Item2Vec. Metode ini mempermudah wisatawan dalam memilih destinasi wisata dan juga mempermudah biro perjalanan dalam menawarkan destinasi wisata untuk keperluan promosi. Metode Item2Vec menghasilkan rekomendasi destinasi wisata di Indonesia dengan akurasi sebesar 61% dan termasuk ke dalam performa cukup baik.
Penerapan Diagram Resetting EWMA Scheme (RES) dalam Mengontrol Coefficient Variation (CV) Kadar Asam Lemak Bebas Bella Novia Syabandina; Suliadi
Bandung Conference Series: Statistics Vol. 3 No. 1 (2023): Bandung Conference Series: Statistics
Publisher : UNISBA Press

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.29313/bcss.v3i1.6998

Abstract

Abstract. One of the tools commonly used in monitoring the production process (Quality Control) is a control chart. The chart that is commonly used is the Shewhart control chart with the assumption that the average (μ) and standard deviation (