IMPLEMENTATION OF DECISION TREE-BASED MACHINE LEARNING ALGORITHM ON WORK FIELD CLASSIFICATION FOR APPRENTICESHIP PARTICIPANTS AT LPK MULIA MANDIRI INDONESIA

• Main Author: Meiko Oke, Yocia
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/55527
• Institution: Institut Teknologi Bandung

Lembaga Pelatihan Kerja Mulia Mandiri Indonesia (LPK MMI) is a company that organizes apprenticeship programs in Japan. Before the apprentices carry out an apprenticeship program in Japan, LPK MMI will first determine the apprentice's field of work based on the apprentice's work experience history. Generally, job applicants' job placement factors are influenced by the knowledge, skills, abilities, preferences, and personality factors of the employee. In fact, the work experience of apprentices is very varied, so, not only does it take a long time to consider all possibilities, but also decision making with only one factor of consideration (the apprentice's work experience factor) does not guarantee a match between the field of work and the apprentice concerned. because of the high possibility of information bias. Therefore, this problem can be a classification problem that can be optimized with machine learning to produce an appropriate field classification for each apprentice based on all the apprentice data held by LPK MMI. Model development will be carried out using a popular classification algorithm, namely the Decision Tree and Random Forest. This final project will use the CRISP-DM methodology which includes the stages of business understanding, data understanding, data preparation, modelling, and evaluation. The main model's performance will be measured from accuracy metrics and weighted F1-score. Overall, Random Forest performs better than the Decision Tree. In addition, the total training duration and predictions required by each algorithm are also measured. Overall the Decision Tree has a much shorter duration than Random Forest.