Search for dissertations about: "usage statistics"

Showing result 1 - 5 of 43 swedish dissertations containing the words usage statistics.

  2. 1. Contributions to Kernel Equating

    Author : Björn Andersson; Fan Yang-Wallentin; Marie Wiberg; Alina A. von Davier; Jorge González; Uppsala universitet; []
    Keywords : NATURVETENSKAP; NATURAL SCIENCES; observed-score test equating; item response theory; R; equipercentile equating; asymptotic standard errors; non-equivalent groups with anchor test design; Statistics; Statistik;

    Abstract : The statistical practice of equating is needed when scores on different versions of the same standardized test are to be compared. This thesis constitutes four contributions to the observed-score equating framework kernel equating. READ MORE

  4. 2. Learning local predictive accuracy for expert evaluation and forecast combination

    Author : Oscar Oelrich; Mattias Villani; Francesco Ravazzolo; Stockholms universitet; []
    Keywords : NATURVETENSKAP; NATURAL SCIENCES; Bayesian; forecast combination; predictive density; Gaussian process; bootstrap; Bayes factors; model selection; Bayesian predictive synthesis; nonparametric methods; power transformation; expected log predictive density; variable selection; statistik; Statistics;

    Abstract : This thesis consists of four papers that study several topics related to expert evaluation and aggregation. Paper I explores the properties of Bayes factors. Bayes factors, which are used for Bayesian hypothesis testing as well as to aggregate models using Bayesian model averaging, are sometimes observed to behave erratically. READ MORE

  5. 3. Regularised Weights in Statistical Models

    Author : Olof Zetterqvist; Chalmers tekniska högskola; []
    Keywords : NATURVETENSKAP; NATURAL SCIENCES; Lasso; Robust Statistics; Deep Learning; Weighted loss.; Noisy Labels; Neural Networks; Regularisation;

    Abstract : For flexible and overparameterised models like neural networks, overfitting can be a notorious problem that makes it hard to give accurate predictions in real-life usage. Overfitting is in particular likely in the presence of errors in the training data, such as misclassifications or outliers. READ MORE

  6. 4. Estimation of wood fibre length distributions from censored mixture data

    Author : Ingrid Svensson; Sara Sjöstedt - de Luna; Lennart Bondesson; Aila Särkkä; Umeå universitet; []
    Keywords : NATURVETENSKAP; NATURAL SCIENCES; censoring; fibre length distribution; identifiability; increment core; length bias; mixture; stochastic EM algorithm; Mathematical statistics; Matematisk statistik;

    Abstract : The motivating forestry background for this thesis is the need for fast, non-destructive, and cost-efficient methods to estimate fibre length distributions in standing trees in order to evaluate the effect of silvicultural methods and breeding programs on fibre length. The usage of increment cores is a commonly used non-destructive sampling method in forestry. READ MORE

  7. 5. Asymptotic Analysis of Machine Learning Models: Comparison Theorems and Universality

    Author : David Bosch; Chalmers tekniska högskola; []
    Keywords : NATURVETENSKAP; NATURAL SCIENCES; NATURVETENSKAP; NATURAL SCIENCES; CGMT; Machine Learning; convex gaussian min max theorem; Asymptotic; universality;

    Abstract : The study of Machine Learning models in asymptotic regimes, has provided insight into many of the properties of ML models, but seemingly contradicts classical statistical wisdom. To solve this mystery, this thesis focuses on the analysis of models such as the LASSO and Random features regression, when the data points and model parameters grow infinite at constant ratios. READ MORE