Danger theory inspired artificial immune system for pattern recognition
Increasing intensive studies on reflecting human immune system's mechanism in computer systems has developed a new computing framework which is called the Artificial Immune System (AIS). Numerous AIS-based applications have been discovered such as anomaly detection system and user-preference b...
Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Research Report |
| Language: | en |
| Published: |
Universiti Malaysia Sabah
2007
|
| Subjects: | |
| Online Access: | https://eprints.ums.edu.my/id/eprint/22875/1/Danger%20theory%20inspired%20artificial%20immune%20system%20for%20pattern%20recognition.pdf https://eprints.ums.edu.my/id/eprint/22875/ |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| Summary: | Increasing intensive studies on reflecting human immune system's mechanism in computer systems has developed a new computing framework which is called the Artificial Immune System (AIS). Numerous AIS-based applications have been discovered such as anomaly detection system and
user-preference based web contents applications of which possess pattern recognition ability. Negative Selection (NS) algorithm is the most commonly researched algorithm within the AIS framework. NS algorithm discriminates 'self' and 'non-self' patterns by acknowledging incoming negative instances. However, as immunology evolves, a theory, the Danger Theory (On, suggested that immune system reacts towards danger signals rather than self and non-self discrimination. Based on ongoing initiatives in outlining OT
algorithm, this research evaluates the performance of OT against NS algorithm within pattern recognition domain.
Perkembangan kajian intensif dalam penafsiran sistem immunisasi
manusia kepada sistem komputer telah melahirkan rangka kerja teknologi
komputer baru yang juga dikenali sebagai Sistem Imun Buatan atau Artirlcial
Immune Systems (AIS). Pelbagai aplikasi berdasarkan AlS telah dipelopori
seperti sistem pengesan kelainan dan kandungan laman sesawang
teradaptasi berdasarkan kecenderungan pengguna yang mana keduanya
mempunyai kebolehan untuk pengesanan pola. Algoritma Pemilihan Negatif
ataupun Negative Selection (NS) adalah antara algoritma yang paling popular
dalam kajian rangka kerja AlS. Algoritma NS mampu membezakan corak data
'diri' dan 'bukan diri' dengan meramal dan mempelajari data negatif.
Walaubagaimanapun, evolusi kajian dalam bidang imunisasi telah
menemukan Teori Bahaya atau Danger Theory (On yang mencadangkan
bahawa sistem imun lebih memberikan reaksi terhadap isyarat bahaya
daripada kebolehan mendiskriminasi 'diri' dan 'bukan diri'. Berdasarkan
penemuan kajian OT yang terkini, penilaian terhadap algoritma OT dan NS
dalam pengesanan pola dijalankan sebagai agenda utama kajian ini. |
|---|