Include featured image in post

georgedouzas · Nov 23, 2023 · d386917 · d386917
1 parent 406e890
commit d386917
Show file tree

Hide file tree

Showing 2 changed files with 2 additions and 1 deletion.
diff --git a/docs/index.html b/docs/index.html
@@ -255,7 +255,7 @@ <h3 class="no-anchor listing-title">
 </a>
 </div>
 </div>
-<div class="quarto-post image-right" data-index="2" data-categories="Machine Learning,Open Source,Imbalanced Data" data-listing-date-sort="1651352400000" data-listing-file-modified-sort="1700745411828" data-listing-date-modified-sort="NaN" data-listing-reading-time-sort="6" data-listing-word-count-sort="1034">
+<div class="quarto-post image-right" data-index="2" data-categories="Machine Learning,Open Source,Imbalanced Data" data-listing-date-sort="1651352400000" data-listing-file-modified-sort="1700745727307" data-listing-date-modified-sort="NaN" data-listing-reading-time-sort="6" data-listing-word-count-sort="1034">
 <div class="thumbnail">
 <p><a href="./projects/machine-learning/geometric-smote/notebooks/index.html" class="no-external"></a></p><a href="./projects/machine-learning/geometric-smote/notebooks/index.html" class="no-external">
 <p><img src="./projects/machine-learning/geometric-smote/notebooks/featured.png" class="thumbnail-image"></p>

diff --git a/docs/projects/machine-learning/geometric-smote/notebooks/index.html b/docs/projects/machine-learning/geometric-smote/notebooks/index.html
@@ -198,6 +198,7 @@ <h1 class="title">Geometric SMOTE algorithm</h1>
 
 
 
+<p><img src="featured.png" class="img-fluid"></p>
 <p>The <a href="https://arxiv.org/pdf/1106.1813.pdf">SMOTE</a> algorithm is the most popular oversampler, with many proposed variants. On the other hand, <a href="https://www.sciencedirect.com/science/article/abs/pii/S0020025519305353">Geometric SMOTE</a> is not another member of the SMOTE’s family since it expands the data generation area and does not use linear interpolation for new samples. You can check the following figure for a visual representation of the their difference:</p>
 <p><img src="smote_vs_gsmote.png" class="img-fluid"></p>
 <p>I have developed a Python implementation of the Geometric SMOTE oversampler called <a href="https://github.com/georgedouzas/geometric-smote">geometric-smote</a>, which integrates with the <a href="https://scikit-learn.org/stable/">Scikit-Learn</a> and <a href="https://imbalanced-learn.org/stable/">Imbalanced-Learn</a> ecosystems. To run a comparison experiment, let’s first create various imbalanced binary datasets with different characteristics:</p>