Claim Missing Document
Check
Articles

Found 3 Documents
Search
Journal : Journal of Computer System and Informatics (JoSYC)

K-Nearest Neighbor (KNN) Algorithm to Determine the Stock of Building Material Store Materials Safitri, Delilla; Fakhriza, Muhammad
Journal of Computer System and Informatics (JoSYC) Vol 5 No 4 (2024): August 2024
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/josyc.v5i4.5731

Abstract

In recent months, a lot of infrastructure has been built, resulting in a shortage of goods in the warehouse due to increased demand for consumer goods and some goods not being sold. Such was the case in January and February 2024 when Riko Jaya panglong experienced a shortage of sand and cement supplies, causing losses. This makes it difficult to predict the inventory of an item in the warehouse. Inventory of goods has great strategic importance for the company. This prediction is very useful in determining the amount of goods to be shipped in the following month. Therefore, companies must implement proactive inventory management. The K-Nearest Neighbor algorithm which looks at the ecluiden distance between old cases and is compared with new cases in an effort to recognize supervised data or data that already exists and has been recorded to help make decisions on the latest cases, this algorithm is very widely applied in other studies because this algorithm has very simple steps and logical reasoning processes by producing the right data and decisions. This data is processed to determine the classification of goods whether increasing or decreasing. And the K-NN algorithm with a value of k = 3 is used to predict stock items. The test results show that K-NN can provide accurate predictions by calculating the Euclidean distance between testing data and training data. The prediction accuracy obtained from the Confusion Matrix reached 100%, indicating the high reliability of this model. Implementation of the K-NN algorithm in RapidMiner with cross-validation technique resulted in a performance of 71.43% for decreasing classification and 67.57% for increasing classification, showing the efficiency of the algorithm in classifying stock data.
Analisis Sentimen Program Makan Gratis Pada Media Sosial X Menggunakan Metode NLP Anggriyani, Wisda; Fakhriza, Muhammad
Journal of Computer System and Informatics (JoSYC) Vol 5 No 4 (2024): August 2024
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/josyc.v5i4.5826

Abstract

This study aims to analyze public sentiment toward the free meal program initiated on Social Media X. Utilizing Natural Language Processing (NLP) methods klasification navie bayes, this research processes text data collected from various user comments and posts on the platform. The collected data is then classified into positive, negative, and neutral sentiment categories. The analysis process involves text preprocessing techniques, including tokenization, stemming, and stop words removal, to enhance the accuracy of the sentiment model. The analysis results show that most users responded positively to the program, particularly regarding the social benefits it offers. However, some negative sentiments were also detected, primarily related to the program's implementation and the quality of the provided meals. These findings offer valuable insights for program organizers to comprehensively understand public perception and make improvements in the future. This study also highlights the importance of using NLP in social media data analysis as a tool to identify and understand public opinion on a large scale.
Penentuan Desa Terbaik dengan Menerapkan Kombinasi Metode ROC dan SMART Margolang, R. Siddik; Fakhriza, Muhammad
Journal of Computer System and Informatics (JoSYC) Vol 5 No 4 (2024): August 2024
Publisher : Forum Kerjasama Pendidikan Tinggi (FKPT)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.47065/josyc.v5i4.5834

Abstract

The problem that is used as a reference in this research is the assessment process which still uses the old method, namely by opening pages of documents, so this method takes a lot of time and energy in the assessment process and allows tolerance of values ​​that are similar, making it very difficult to determine the best village. The process of selecting the best village is a very important stage in building a competent and innovative village. The system built using the ROC and SMART methods aims to help data collectors reduce time in data collection and be effective in determining the best villages according to the criteria used. The criteria used are Village Cleanliness (K1), Education (K2), Economy (K3), Security and Order (K4), Role of the PKK (K5), Village Governance (K6) with 7 alternative data samples. Data for this research was collected through interviews and observations at the Rahuning District Office and processed using the ROC and SMART methods for weighting and evaluating decisions. The results of calculations using the ROC and SMART methods found that among the best villages, Gunung Melayu Village was a worthy candidate to become the best village candidate by obtaining a total score of 0.738889.