Peningkatan Akurasi Klasifikasi Tingkat Agility Menggunakan Metode Fuzzy-Particle Swarm Optimization

Nugroho, Tri Yulianto (2025) Peningkatan Akurasi Klasifikasi Tingkat Agility Menggunakan Metode Fuzzy-Particle Swarm Optimization. Masters thesis, Institut Teknologi Sepuluh Nopember.

Text 6025222015-Master_Thesis.pdf - Accepted Version Restricted to Repository staff only until 1 April 2027. Download (2MB) | Request a copy

Abstract

Suksesnya proses pengembangan perangkat lunak bergantung pada komposisi tim pengembangan yang tepat. Menurut Agile Manifesto(2001) Statement of Values, kualitas individu dan tim merupakan salah satu variabel paling utama dalam menentukan tingkat keberhasilan pengembangan produk perangkat lunak berbasis Agile. Sayangnya, banyak pengembang perangkat lunak belum menyadarinya. Kualitas tim yang baik ditandai dengan meningkatnya agility (ketangkasan). Sebaliknya, terjadinya praktik-praktik buruk berupa anti patterns dalam tim akan menurunkan kualitasnya. Dalam penelitian ini, akan dilakukan pengukuran terhadap berbagai aspekpada anggota tim pengembang, baik secara individu maupun tim. Data asesmen meliputi kualifikasi anggota tim pengembang dari segi hard skill maupunsoft skilluntuk mengukur tingkat agility. Selanjutnya, dengan menerapkan metode fuzzy particle swarm optimization, dilakukan optimasi klasifikasi untukmeningkatkanhasil agility. Melalui antarmuka berbasis web, pengguna memasukkan komposisi tim yang hendak direncanakan dalam proyek. Sistemakan memberikan keluaran berupa tingkat agility yang akan dirancang. Kontribusi utama penelitian ini adalah menyediakan alat ukur tingkat agility berdasarkan antipattern,Keirsey dan peran. Obyek penelitianini adalahmahasiswa program D4 Teknik Informatika PENS di semester 4 dan5. Evaluasidilakukan menggunakan metode DecisionTreeClassifier train test split 80:20untukmemvalidasi model yang dibangun. Hasil akhir membuktikan bahwa metode gabungan Fuzzy-PSO dpat ngklasifikasikan tingkat agility lebih optimal dibandingkan dengan mtode ndom Forest atau KNN, dengan tingkat akurasi rata-rata mencapai 90.6% dngan ta-rata waktu komputasi 0.37 detik pada 160 data peserta. Penelitian ii harapkan dapat menghasilkan sebuah pustaka untuk mengklasifikasikan pribadian dan peran dalam tim secara optimal.
==================================================================================================================================
The success of the software development process depends ontheproper composition of the development team. According to the AgileManifesto(2001) Statement of Values, the quality of individuals and teams is oneofthemost crucial variables in determining the success of Agile-basedsoftwareproduct development. Unfortunately, many software developers haveyet torealize this. A high-quality team is characterized by an increase inagility. Conversely, poor practices in the form of antipatterns within theteamcandegrade its quality. This study measures various aspects of development teammembers, both individually and as a team. Assessment data includes the qualificationsofteam members in terms of hard skills and soft skills to evaluate thelevel ofagility. Subsequently, by applying the fuzzy particle swarmoptimizationmethod, classification optimization is performed to enhance agilityoutcomes. Through a web-based interface, users can input the compositionoftheteam planned for a project. The system will output the designed agilitylevel. The primary contribution of this research is to provide a tol for asuring agility levels based on antipatterns, Keirsey's personality tpes, and les. The research subjects consist of 4th and 5th-semester students of the D formatics Engineering program at PENS. The evaluation is conducted uing e DecisionTreeClassifier method with an 80:20 train-test split to vlidate te del. The results demonstrate that the combined Fuzzy-PSO mthod cn assify agility levels more optimally compared to the Random Frest or KNN thods, achieving an average accuracy of 90.6% wth an average computation me of 0.37 seconds on 160 participant data samples. This study i expected t oduce a framework for optimally classifying personalities and rles within a am.

Item Type:	Thesis (Masters)
Uncontrolled Keywords:	MBTI, Keirsey, Fuzzy, PSO, AntiPola, Particle-Swarm- Optimiization, pengembangan perangkat lunak, Agility, Scrum
Subjects:	T Technology > TA Engineering (General). Civil engineering (General) T Technology > TA Engineering (General). Civil engineering (General) > TA660.F7 Structural frames.
Divisions:	Faculty of Intelligent Electrical and Informatics Technology (ELECTICS) > Informatics Engineering > 55101-(S2) Master Thesis
Depositing User:	Tri Yulianto Nugroho
Date Deposited:	01 Feb 2025 06:30
Last Modified:	01 Feb 2025 06:30
URI:	http://repository.its.ac.id/id/eprint/117340

Actions (login required)

View Item