2024
An entropy-based class of moving averages
Andreas Kull. Journal of Investment Strategies, 13(1), 1–13.
Maximum-entropy kernels unify familiar moving averages and a broader class of backward filters.
View paperFinite protocols for inference and measurement.
About
Apeirics is a consulting and research firm specialising in quantitative analysis and strategic risk advisory.
Our domain of expertise is inference in high-dimensional and unbounded parameter spaces. We combine maximum-entropy reasoning with finite measurement to determine what follows from the available constraints — and to establish bounds where point estimates would claim more than the evidence supports.
Our foundational focus is finite measurement protocols that give rise to intrinsic fingerprints of the recording process, beyond what is measured.
Research
Selected work on maximum-entropy inference and finite measurement protocols.
Maximum entropy
2024
Andreas Kull. Journal of Investment Strategies, 13(1), 1–13.
Maximum-entropy kernels unify familiar moving averages and a broader class of backward filters.
View paper2022
Andreas Kull. Journal of Investment Strategies, 11(2), 47–58.
An information-theoretic derivation of least-biased investment strategies and their link to the Kelly criterion.
View paperFinite measurement
Forthcoming
Andreas Kull.
White papers
2003
Andreas Kull. Casualty Actuarial Society Forum, Winter 2003, 317–350.
Maximum entropy selects a martingale measure that connects insurance and financial risk pricing.
Read white paperContact
Regulatory & Legal
This website is hosted by Infomaniak in Switzerland. Technical server logs may be processed to operate and secure the site. We use no third-party analytics, advertising trackers or non-essential cookies. Email data is used only to handle enquiries and related business and retained only as necessary or required by law.
Site content is general information, not financial, investment, legal or other professional advice; no client relationship arises from its use. No warranty is made as to completeness, accuracy or availability, and liability is excluded to the extent permitted by law.