DETERMINATION OF ALUMNI WORK ABSORPTION TYPE WITH C4.5 CLASSIFICATION ALGORITHM CASE STUDY: FACULTY OF COMPUTER SCIENCE, MERCU BUANA UNIVERSITY

BAHRI, MUHAMAD ZAINUL (2019) DETERMINATION OF ALUMNI WORK ABSORPTION TYPE WITH C4.5 CLASSIFICATION ALGORITHM CASE STUDY: FACULTY OF COMPUTER SCIENCE, MERCU BUANA UNIVERSITY. S1 thesis, Universitas Mercu Buana Jakarta.

Text (JURNAL MAHASISWA) Laporan Yudisium pake M peprus.pdf Restricted to Registered users only Download (1MB)

Abstract

According to the Central Bureau of Statistics in 2018, the number of unemployed people among the University is 789,113 people and this number is increasing annually, so that continuous improvement is needed. The university as an institution of higher education consider it necessary to do an alumni type mapping based on work waiting time, this is due to increasing unemployment from the university. The intended mapping is fast, medium, and slow to get a job. To calcify the type of alumni needs to use a data-mining approach with the C4.5 algorithm and implement in RapidMiner tools in order to see the results of the classification, namely fast, medium, and slow to get a job. The results of the confusion matrix calculation produce near-perfect accuracy of 98.78% and kappa 0.935. From the results of the application of the C4.5 algorithm it can be concluded that predicting alumni with a waiting time to work based on their GPA values is very satisfying and is expected to help the Community of Mercu Buana to increase Mercu Buana University graduates to get a job. Key words: Data Mining, Classification, Decision Tree, C4.5 Algorithm, Waiting Time Work. Menurut Badan Pusat Statistik tahun 2018 jumlah penganggur dikalangan Universitas sebesar 789,113 Jiwa jumlah ini meningkat setiap tahunnya. Sehingga diperlukan perbaikan yang berkelanjutan, universitas sebagai institusi pendidikan tinggi memandang perlu untuk melakukan sebuah pemetaan tipe alumni berdasarkan waktu tunggu bekerja, hal ini dikarenakan makin banyaknya pengangguran dari kalangan universitas. pemetaan yang dimaksud seperti cepat,sedang dan lambat untuk mendapatkan pekerjaan. Untuk melakukan kalsifikasi tipe alumni perlu menggunakan pendekatan berbasis data mining dengan algoritma C4.5 dan diimplementasikan ke tools rapid miner supaya dapat melihat hasil dari klasifikasi tersebut yaitu cepat,sedang dan lambat mendapatkan pekerjaan. Hasil perhitungan confusion matrix menghasilkan akurasi hampir sempurna yaitu 98.78% dan kappa 0.935. Dari hasil penerapan algoritma C4.5 dapat disimpulkan bahwa memprediksi alumni dengan waktu tunggu bekerja berdasarkan nilai IPK akurasinya sangat memuaskan dan diharapkan dapat membantu Civitas Mercu buana untuk meningkatkan lulusan Universitas Mercu buana untuk mendapatkan pekerjaan. Kata kunci: Data mining, Klasifikasi, Decision tree, Algoritma C4.5, Waktu tunggu bekerja

Actions (login required)

View Item