OleksandrZadorozhnyiML/StMaRDI: Structure learning notebooks for MaRDI TA3 v3. (Q6241082): Difference between revisions

From MaRDI portal
Added link to MaRDI item.
Removed claim: author (P16): Item:Q6241080
Property / author
Q6241080 (Deleted Item)
 
Property / author: Q6241080 (Deleted Item) / rank
Normal rank
 

Revision as of 09:17, 16 July 2024

Software published at Zenodo repository
Language Label Description Also known as
English
OleksandrZadorozhnyiML/StMaRDI: Structure learning notebooks for MaRDI TA3 v3.
Software published at Zenodo repository

    Statements

    0 references
    <p>This is a new release which stores the notebooks of statistical analysis in graphical modelling and causal inference (MaRDI Project, TA3)</p><p>Here we provide the short description of the available notebooks (5 notebooks are avalaible currently).</p><p><strong>Notebook_01</strong> - introduction to the problem of graphical modelling and causal inference. Implementation of the algorithms available in the package bnlearn on the dataset alarm which were collected from Zenodo repository.</p><p><strong>Notebook_02</strong> - introduction to the problem of subgraph (subset) selection in a &quot;large&quot; graph with available ground truth such that the independence relationships in the selected graph do not contradict to the ones in the supergraph. Implementation of different heuristics based on alarm graph and their benchmarking.</p><p><strong>Notebook_03</strong> - based on the data collected by telekom company estimate which covariate cause clients to become churn. Also a comparison to known methods which provide feature importance is presented. This notebook can also serve as an introductory notebook in the courses of statistical analysis.</p><p><strong>Notebook_04</strong> - benchmarking datasets from Tuebingen cause-effect repository by means of a statistical test which uses kernel-based HSIC methodology.</p><p><strong>Notebook_05</strong> - a notebook which introduces the problem of DAG estimation as a non-convex optimization problem over euclidean space. Implementation uses dataset from Bayesys collection collected and published earlier in zenodo community.</p>
    0 references
    6 February 2024
    0 references

    Identifiers

    0 references