Claim Missing Document
Check
Articles

Found 12 Documents
Search

Fire Image Set for Evoking Panic sabilirrasyad, Iqbal; Basuki, Achmad; Harsono, Tri
EMITTER International Journal of Engineering Technology Vol 8 No 2 (2020)
Publisher : Politeknik Elektronika Negeri Surabaya (PENS)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.24003/emitter.v8i2.504

Abstract

Fire is the closest disaster to us, a person who put cigarettes around flammable objects could burn one to dozens of houses. The last thing that happens was a mass panic. In this kind of situation, panic is one of the keys to determine how much probability someone will survive. However, detecting someone's panic during a fire is impossible. This leads some scientists to assume that mass panic was never happening and some studies use simple functions to determine someone when panic. Currently, thanks to technological advances we can easily build virtual worlds that resemble real events. To build a virtual world that could evoke panic we still need the right stimulus. In this paper, we will discuss with terms of fire disaster stimulus that possible to impel someone to feel panic. While some stimulus datasets that already exist have more broad categories, we wanted to focus on a specific problem. The determined parameters are considered through several elements that could cause a person to panic, either before or during a fire. By using the Self-Assessment Manikin system to obtain valance and arousal matrix, we conduct a test to see how much influence the fire categories stimulus provided.
DETEKSI ANOMALI VOLUME DAN HARGA PADA SAHAM ASURANSI IDX MENGGUNAKAN ISOLATION FOREST Iqbal Sabilirrasyad
JOURNAL OF SCIENCE AND SOCIAL RESEARCH Vol. 9 No. 4 (2026): August 2026 (1)
Publisher : Smart Education

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.54314/jssr.v9i4.7033

Abstract

Asuransi Indonesia pernah alami kegagalan besar (Jiwasraya 2020-an, Prolife diinvestigasi). Hal itu mempertegas pentingnya pengawasan ketat pada emiten asuransi di BEI, terutama yang berkapitalisasi kecil dan likuiditas timpang. Penelitian ini menerapkan Isolation Forest untuk mendeteksi anomali volume dan harga saham sektor asuransi IDX menggunakan data harian End-of-Day periode 2 Januari 2020 hingga 29 Mei 2026. Audit likuiditas meloloskan 14 dari 18 saham (77,8%) ke tahap pemodelan; 4 saham (LPGI, ASRM, LIFE, ABDA) dikeluarkan karena proporsi hari tanpa transaksi berlebihan. Tiga fitur, yaitu log return, rasio volume relatif terhadap MA-20, dan rentang intraday relatif, melatih model per saham dengan ambang anomali pada persentil ke-5; fitur keempat yang direncanakan, yaitu rasio ukuran transaksi, tidak dapat dihitung karena keterbatasan data. Model mengidentifikasi 1.036 hari anomali (rata-rata 5,0% per saham, akibat konstruksi ambang batas, bukan temuan independen). Triangulasi terhadap kalender peristiwa hanya mengonfirmasi 42 anomali (4,1%), sementara enam anomali paling ekstrem tidak berkaitan dengan peristiwa mana pun. Anomali didominasi pola idiosyncratic (97,5% dari hari beranomali) dibanding sector-wide (2,5%), dengan satu hari ekstrem (5 Januari 2026) mencatat keserentakan penuh pada seluruh 14 saham yang memerlukan verifikasi lanjutan. Temuan ini mendukung kelayakan Isolation Forest sebagai alat pemantauan awal bagi investor dan regulator, dengan batasan metodologis yang ditandai eksplisit. Anomalies were dominated by idiosyncratic patterns (97.5% of anomalous days) rather than sector-wide patterns (2.5%), with one extreme day (January 5, 2026) showing complete synchrony across all 14 stocks, which requires further verification. These findings support the viability of Isolation Forest as an early-warning monitoring tool for investors and regulators, with its methodological limitations explicitly noted.