Analisis Kinerja Aparatur Sipil Negara Pemerintah Provinsi Sumatera Utara berdasarkan Nilai Indeks Profesionalitas menggunakan Metode Machine Learning

Abstract

The assessment of the performance of state civil apparatuses is of critical importance for the assurance of effective governance. This research project is concerned with the evaluation of civil servants within the Indonesian bureaucracy, with a particular focus on the North Sumatra provincial government. The primary metric employed in this evaluation is the professionalism index. To enhance the objectivity and efficiency of this assessment, machine learning methods are employed to derive meaningful patterns from extensive datasets. . This research, entitled "Performance Analysis of North Sumatra Province Civil Servant based on Professionalism Index Values using Machine Learning Methods," employs inferential statistics and a range of machine learning algorithms using of the pycaret library, to identify patterns and correlations within a substantial dataset of civil servant performance metrics. Classification algorithms are employed to categorize civil servants based on their level of professionalism, while inferential statistics are utilized to draw broader conclusions about the population of civil servants based on the sample data analyzed. This statistical approach allows for the identification of statistically significant differences and relationships between variables, thereby providing valuable insights into the factors influencing performance. The results of the research indicated a positive correlation between qualifications, competencies, and performance; however, no significant correlation was observed between qualifications and discipline. Regression analysis demonstrated that qualifications, competencies, and performance had a positive effect on the Professionalism Index (IP), with competencies exerting the most significant influence. The machine learning model identified competency as the most crucial dimension in predicting IP, followed by performance and qualifications, while discipline had a negligible impact. Furthermore, the model underscored the importance of other factors, including the type of position, education level, individual performance, and work environment, in shaping ASN professionalism.