IMPLEMENTASI ARTIFICIAL INTELLIGENCE PADA PERMAINAN SNAKE DENGAN METODE REINFORCEMENT LEARNING

Rosyid, Sulistyawan Abdillah (2025) IMPLEMENTASI ARTIFICIAL INTELLIGENCE PADA PERMAINAN SNAKE DENGAN METODE REINFORCEMENT LEARNING. Bachelor thesis, Institut Teknologi Kalimantan.

	Text 11181079_cover.pdf Restricted to Repository staff only until January 2027. Download (279kB) | Request a copy
	Text 11181079_statement_of_authenticity.pdf Restricted to Repository staff only until January 2027. Download (550kB) | Request a copy
	Text 11181079_publishing_agreement.pdf Restricted to Repository staff only until January 2027. Download (1MB) | Request a copy
	Text 11181079_approval_sheet.pdf Restricted to Repository staff only until January 2027. Download (929kB) | Request a copy
	Text 11181079_preface.pdf Restricted to Repository staff only until January 2027. Download (1MB) | Request a copy
	Text 11181079_abstract_id.pdf Restricted to Repository staff only until January 2027. Download (1MB) | Request a copy
	Text 11181079_abstract_en.pdf Restricted to Repository staff only until January 2027. Download (1MB) | Request a copy
	Text 11181079_table_of_content.pdf Restricted to Repository staff only until January 2027. Download (3MB) | Request a copy
	Text 11181079_illustrations.pdf Restricted to Repository staff only until January 2027. Download (844kB) | Request a copy
	Text 11181079_tables.pdf Restricted to Repository staff only until January 2027. Download (499kB) | Request a copy
	Text 11181079_conclusions.pdf Restricted to Repository staff only until January 2027. Download (3MB) | Request a copy
	Text 11181079_bibliography.pdf Restricted to Repository staff only until January 2027. Download (2MB) | Request a copy
	Text 11181079_paper.pdf Restricted to Repository staff only until January 2027. Download (638kB) | Request a copy
	Text 11181079_Form. TA-20.pdf Restricted to Repository staff only until January 2027. Download (690kB) | Request a copy
	Text 11181079_chapter_1.pdf Restricted to Repository staff only until January 2027. Download (1MB) | Request a copy
	Text 11181079_chapter_2.pdf Restricted to Repository staff only until January 2027. Download (2MB) | Request a copy
	Text 11181079_chapter_3.pdf Restricted to Repository staff only until January 2027. Download (1MB) | Request a copy
	Text 11181079_chapter_4.pdf Restricted to Repository staff only until January 2027. Download (2MB) | Request a copy
	Text 11181079_enclosure.pdf Restricted to Repository staff only until January 2027. Download (2MB) | Request a copy
	Text 11181079_presentation.pdf Restricted to Repository staff only until January 2027. Download (1MB) | Request a copy

Abstract

Permainan Snake dipilih sebagai media implementasi Reinforcement Learning (RL) karena kompleksitasnya yang cukup untuk menguji kemampuan agen Artificial Intelligence (AI). Penelitian ini bertujuan mengembangkan agen AI menggunakan algoritma Deep Q-Network (DQN) agar dapat bermain secara optimal. Permasalahan yang dikaji meliputi efektivitas strategi pembelajaran serta pengaruh parameter pelatihan terhadap performa agen. Lingkungan simulasi dibangun menggunakan PyGame dan pelatihan dilakukan selama lebih dari 4000 episode dengan dua pendekatan: Bellman dan Monte Carlo. Hasil menunjukkan bahwa pendekatan Bellman menghasilkan skor rata-rata 23,58 dan panjang ular maksimum 70 unit, sementara pendekatan Monte Carlo menunjukkan skor rata-rata 21–22 dan panjang maksimum 67 unit. Agen mampu menunjukkan perilaku adaptif, seperti menghindari rintangan dan mendekati makanan secara efisien. Temuan ini membuktikan bahwa algoritma DQN mampu membentuk agen yang stabil dan responsif dalam lingkungan permainan yang dinamis.

Item Type:	Thesis (Bachelor)
Subjects:	T Technology > T Technology (General)
Divisions:	Jurusan Matematika dan Teknologi Informasi > Informatika
Depositing User:	Sulistyawan Abdillah Rosyid
Date Deposited:	15 Jul 2025 07:02
Last Modified:	15 Jul 2025 07:02
URI:	http://repository.itk.ac.id/id/eprint/24535

Actions (login required)

View Item