ScholarGate
Asisten
Process / pipelineDeneysel desen

Uji A/B Adaptif — Pengujian A/B Adaptif

Uji A/B Adaptif adalah desain eksperimental yang secara dinamis mengalokasikan ulang lalu lintas atau partisipan ke varian yang berkinerja lebih baik selama eksperimen itu sendiri, alih-alih mempertahankan alokasi tetap hingga akhir. Mengambil dari algoritma bandit multi-lengan seperti Thompson Sampling atau Upper Confidence Bound (UCB), ia menyeimbangkan eksplorasi varian yang tidak pasti dengan eksploitasi varian yang sudah menunjukkan kinerja unggul, biasanya menghasilkan hasil agregat yang lebih tinggi sambil tetap menghasilkan kesimpulan inferensial yang valid.

Temukan Topik dengan PaperMindSegeraVideoSegeraDownload slides

Baca metode selengkapnya

Khusus anggota

Masuk dengan akun gratis untuk membaca bagian ini.

Masuk

Method map

The neighbourhood of related methods — select a node to explore.

Sumber

  1. Russo, D., Van Roy, B., Kazerouni, A., Osband, I., & Wen, Z. (2018). A Tutorial on Thompson Sampling. Foundations and Trends in Machine Learning, 11(1), 1–96. DOI: 10.1561/2200000070
  2. Offer-Westort, M., Coppock, A., & Green, D. P. (2021). Adaptive Experimental Design: Prospects and Applications in Political Science. American Journal of Political Science, 65(4), 826–844. DOI: 10.1111/ajps.12597

Cara menyitasi halaman ini

ScholarGate. (2026, June 3). Adaptive A/B Testing. ScholarGate. https://scholargate.app/id/experimental-design/adaptive-ab-test

Which method?

Set this method beside its closest kin and read them side by side — the library lays the books on the table; the choice is yours.

Compare side by side

Dirujuk oleh

ScholarGateAdaptive A/B test (Adaptive A/B Testing). Diakses 2026-06-15 dari https://scholargate.app/id/experimental-design/adaptive-ab-test · Set data: https://doi.org/10.5281/zenodo.20539026