home.social

#outliers — Public Fediverse posts

Live and recent posts from across the Fediverse tagged #outliers, aggregated by home.social.

  1. 'Outlier Robust and Sparse Estimation of Linear Regression Coefficients', by Takeyuki Sasai, Hironori Fujisawa.

    jmlr.org/papers/v26/23-1583.ht

    #outlier #outliers #robust

  2. #Day7 of #30DayChartChallenge, theme: #Outliers

    📦📉 Comparativa Boxplot: Rentabilidades log diarias de IBEX 35, S&P 500, Nasdaq (~3 años).

    Nasdaq destaca por mayor volatilidad (IQR ~1.64% vs ~1.1%) y outlier máximo (+7.1% el 10-Nov-2022). Todos muestran "colas anchas".

    Data: Yahoo Finance via #rstats {quantmod}. Viz: #ggplot2

    📂 Código: t.ly/JxcM1

    #dataviz #finance #StockMarket #Volatility #quant #IBEX35 #SP500 #Nasdaq

  3. 'PyGOD: A Python Library for Graph Outlier Detection', by Kay Liu, Yingtong Dou, Xueying Ding, Xiyang Hu, Ruitong Zhang, Hao Peng, Lichao Sun, Philip S. Yu.

    jmlr.org/papers/v25/23-0963.ht

    #outliers #outlier #pygod

  4. RT by @EU_ScienceHub: "#AI always pushes out #outliers, which is anyone who’s not a white middle-aged man. So anyone in the audience right now who’s not a white middle-aged man is being treated #unfairly by AI systems.

    That is the rule that we find when we #audit AI systems” - @gemmagaldon at #ESPAS

    🐦🔗: nitter.cz/Eticas_tech/status/1

    [2023-11-14 13:57 UTC]

  5. 'Outlier-Robust Subsampling Techniques for Persistent Homology', by Bernadette J. Stolz.

    jmlr.org/papers/v24/21-1526.ht

    #homology #outliers #topological

  6. Recent psych methods papers:

    Here’s a thread of five helpful articles from 2022 on psychological research methodology. They address outlier exclusion, longitudinal research, interventions, informed consent, and control variables. 🧵👉

    #Outliers
    #Longitudinal
    #Interventions
    #InformedConsent
    #Debriefing
    #ControlVariables
    #Psychology
    #Methodology
    @psychology

    1/6

  7. Food for thought #Samples from higher #paleolatitudes were on average smaller, but contained #outliers with the largest #sizes, highlighting the need to analyze entire #distributions. #Lithology and #collection differences need to be accounted for when evaluating #size differences across #space (#polar #gigantism) and #time (#Lilliput #effect). Similarly, differences in #facies may affect #species #determination

    doi.org/10.2110/palo.2021.063

    #Paleobiology #Paleontology #Ecology #EvolutionaryBiology