@phdthesis{eprintuntirta52836,
            year = {2025},
          author = {A'IDAH EKA SEPTIANA},
          school = {Fakultas Teknik Universitas Sultan Ageng Tirtayasa},
           title = {OPTIMALISASI PROSES KADERISASI DI PERUSAHAAN XYZ 
DENGAN MEMAHAMI PERSEBARAN DATA KARYAWAN 
MENGGUNAKAN ALGORITMA FUZZY C-MEANS},
            note = {Proses kaderisasi merupakan aspek utama dalam pengembangan Sumber Daya Manusia (SDM) di perusahaan. Penelitian ini bertujuan untuk mengelompokkan karyawan berdasarkan karakteristik tertentu dengan mengimplementasikan algoritma Fuzzy C-Means (FCM) guna mendukung pengambilan keputusan berbasis data dalam proses kaderisasi. FCM dipilih karena kemampuannya dalam menangani data yang memiliki kemiripan antar cluster melalui derajat keanggotaan. Dataset yang digunakan mencakup variabel department, region, education, gender, recruitment channel, no of trainings, age, previous year rating, length of service, KPIs  met more than 80, awards won, dan average training score. Pendekatan utama dalam penelitian ini menggunakan seluruh fitur dalam analisis clustering agar informasi yang didapatkan mencakup seluruh informasi yang relevan dengan karyawan. Sebagai pembanding, dilakukan juga analisis menggunakan feature selection yang menunjukkan peningkatan nilai evaluasi cluster, ditunjukkan dengan meningkatnya nilai Silhouette Score dari 0.7934 menjadi 0.8634 dan Fuzzy Partition Coefficient (FPC) dari 0.8918  menjadi 0.9347, serta penurunan Xie-Beni Index (XBI) dari 0.0222 menjadi 0.0194. Meskipun demikian, penggunaan feature selection mengurangi keluasan informasi yang diperoleh dari hasil clustering karena terbatasnya fitur yang digunakan, sehingga perlu dipertimbangkan dalam mengimplementasikannya. Berdasarkan hasil akhir clustering dengan menggunakan seluruh fitur, didapatkan cluster optimal saat berjumlah 4 cluster yang merepresentasikan cluster karyawan dengan karakteristik yang berbeda. Hal tersebut dapat digunakan dalam menyusun strategi kaderisasi yang tepat, efektif, dan berbasis data.},
           month = {July},
             url = {https://eprints.untirta.ac.id/52836/},
        abstract = {The cadre development process is a key aspect of human resource (HR) development in companies. This study aims to group employees based on specific characteristics by 
implementing the Fuzzy C-Means (FCM) algorithm to support data-driven decision making in the cadre development process. FCM was chosen for its ability to handle data with similarities between clusters through membership degrees. The dataset used includes variables such as department, region, education, gender, recruitment channel, number of trainings, age, previous year rating, length of service, KPIs met more than 80, awards won, and average training score. The main approach in this study uses all features in the clustering analysis to ensure that the information obtained covers all information relevant to employees. As a comparison, an analysis using feature selection was also conducted, showing an improvement in cluster evaluation values, as indicated by an increase in the Silhouette Score from 0.7934 to 0.8634 and the Fuzzy Partition Coefficient (FPC) from 0.8918 to 0. 9347, as well as a decrease in the Xie-Beni Index (XBI) from 0.0222 to 0.0194. However, the use of feature selection reduces the breadth of information obtained from the clustering results due to the limited features used, so this should be considered when implementing it. Based on the final clustering results using all features, the optimal number of clusters was found to be 4, representing employee clusters with distinct characteristics. This can be used to develop appropriate, effective, and data-driven employee development strategies.},
        keywords = {Kaderisasi Karyawan, Fuzzy C-Means, Silhouette Score, Fuzzy Partition Coefficient, Xie-Beni Index}
}