No abstract

Pada industri jasa pelayanan peti kemas, Terminal Nilam merupakan pelanggan dari PT. BIMA, yang secara khusus bergerak dibidang jasa perbaikan dan perawatan alat berat. Terminal ini menjadi sentral tempat untuk melakukan aktifitas bongkar muat peti kemas domestik yang memiliki empat buah container crane untuk melayani dua kapal. Proses perawatan alat berat seperti container crane yang selama ini beroperasi, agaknya kurang memperhatikan data pengelompokkan atau klasifikasi jenis perawatan yang dibutuhkan oleh alat berat tersebut. Di kemudian hari, alat berat dapat menunjukkan kinerja yang tidak maksimal bahkan dapat berujung pada kecelakaan kerja. Selain itu, kelalaian perawatan container crane juga dapat menyebabkan pembengkakan biaya perawatan lanjut. Target produksi bongkar muat dapat berkurang dan juga keterlambatan jadwal kapal sandar sangat mungkin terjadi. Metode pembelajaran menggunakan mesin atau biasa disebut dengan Machine Learning (ML), dengan mudah dapat melenyapkan kemungkinan-kemungkinan tersebut. ML dalam penelitian ini, kami rancang agar bekerja dengan mengidentifikasi lalu mengelompokkan jenis perawatan container crane yang sesuai, yaitu ringan atau berat. Metode ML yang pilih untuk digunakan dalam penelitian ini yaitu Random Forest, Support Vector Machine, k-Nearest Neighbor, Naïve Bayes, Logistic Regression, J48, dan Decision Tree. Penelitian ini menunjukkan keberhasilan ML model tree dalam melakukan pembelajaran jenis data perawatan container crane (numerik dan kategoris), dengan J48 menunjukkan performa terbaik dengan nilai akurasi dan nilai ROC-AUC mencapai 99,1%. Pertimbangan klasifikasi kami lakukan dengan mengacu kepada tanggal terakhir perawatan, hour meter, breakdown, shutdown, dan sparepart.

<span lang="EN-US">This study describes the methods for eliminating irrelevant features in speech data to enhance toxic speech classification accuracy and reduce the complexity of the learning process. Therefore, the wrapper method is introduced to estimate the forward selection technique based on support vector machine (SVM) and random forest (RF) classifier algorithms. Eight main speech features were then extracted with derivatives consisting of 9 statistical sub-features from 72 features in the extraction process. Furthermore, Python is used to implement the classifier algorithm of 2,000 toxic data collected through the world's largest video sharing media, known as YouTube. Conclusively, this experiment shows that after the feature selection process, the classification performance using SVM and RF algorithms increases to an excellent extent. We were able to select 10 speech features out of 72 original feature sets using the forward feature selection method, with 99.5% classification accuracy using RF and 99.2% using SVM.</span>

No abstract

This study aims to optimize servers with low utility levels on hardware using container virtualization techniques from Docker. This study's primary focus is to maximize the work of the CPU, RAM, and Hard Drive. The application of virtualization techniques is to create many containers as each of the containers is for the application to run a cloud storage system with the CaaS service infrastructure concept (Container as a Service). Containers on infrastructure will interact with other containers using configuration commands at Docker to form an infrastructure service such as CaaS in general. Testing of hardware carried out by running five Nextcloud cloud storage applications and five MariaDB database applications running in Docker containers and tested by random testing using a multimedia dataset. Random testing with datasets includes uploading and downloading datasets simultaneously and CPU monitoring under load, RAM, and Disk hardware resources. The testing will be done using Docker stats, HTOP, and Cockpit monitoring tools to determine the hardware capabilities when processing multimedia datasets.