From 3e49ea19f839b847e71610638d0c212ea2292925 Mon Sep 17 00:00:00 2001
From: unknown <cristianvergaraf@hotmail.com>
Date: Fri, 6 Sep 2024 12:50:35 +0200
Subject: [PATCH] adding links

---
 _posts/2024-08-30-land-use-classification.md.md | 12 ++++++------
 1 file changed, 6 insertions(+), 6 deletions(-)

diff --git a/_posts/2024-08-30-land-use-classification.md.md b/_posts/2024-08-30-land-use-classification.md.md
index c3ff399..fa9db9c 100644
--- a/_posts/2024-08-30-land-use-classification.md.md
+++ b/_posts/2024-08-30-land-use-classification.md.md
@@ -12,7 +12,7 @@ title: Land use classification of Landsat images
 <h3 style = "font-weight:800;"> Transferable Technical Skills Applied in This Project
 </h3>
 
-* [Data extraction] (JavaScript, Google Earth Engine)
+* [Data extraction](https://github.com/cristianvergaraf/land_use_classification/blob/main/download_images.js) (JavaScript, Google Earth Engine)
 * Generation of training set (QGIS, Python)
 * Data preprocessing and exploratory analysis (Pandas)
 * Testing different machine learning algorithms: SVM, Random Forest (scikit-learn)
@@ -28,24 +28,24 @@ In this example the Lingue basin in south-central Chile was classified into seve
 ![land use cover](land_use_web.jpg)
 
 
-<h3 style = "font-weight:800;"> Data Extraction</h3>
+<h3 style = "font-weight:800;"> <a href = "https://github.com/cristianvergaraf/land_use_classification/blob/main/download_images.js" > Data Extraction </a> </h3>
 
 Download images from Google Earth Engine using simple javaScript.
 
-<h3 style = "font-weight:800;"> Generation of Training Sites</h3>
+<h3 style = "font-weight:800;"> <a href = "https://github.com/cristianvergaraf/land_use_classification/blob/main/extract_data.ipynb"> Generation of Training Sites </a></h3>
 
 A training set  was generated by selecting sites through photointerpretation of high-resolution images (Google Earth). The selected classes were natural forest, forest plantation, shrubland, agriculture, urban, and water. 
 
-<h3 style = "font-weight:800;"> Exploratory Analysis</h3>
+<h3 style = "font-weight:800;"> <a href = "https://github.com/cristianvergaraf/land_use_classification/blob/main/exploratory_analysis_final.ipynb">Exploratory Analysis</a></h3>
 
 Since the reflectance values were already scaled from 0 to 1, additional scaling was unnecessary. The distribution of bandas, separability of classes, and identification of outliers were performed to assess the training sites.
 
-<h3 style = "font-weight:800;"> Testing Different Machine Learning Algorithms</h3>
+<h3 style = "font-weight:800;"> <a href = "https://github.com/cristianvergaraf/land_use_classification/blob/main/machine_learning_final.ipynb"> Testing Different Machine Learning Algorithms</a></h3>
 
 Three different machine algorithms were compared to determine which performed best for this task, considering the small and imbalanced number of training sites.
 
 
-<h3 style = "font-weight:800;"> Predicting Land Use Classes</h3>
+<h3 style = "font-weight:800;"> <a href = https://github.com/cristianvergaraf/land_use_classification/blob/main/clasificacion_imagen.ipynb> Predicting Land Use Classes </a></h3>
 
 The models were applied to predict the land use/cover classes for the entire image. Rasterio was used to manage the image and transform it into an array, enabling model application. The results were then converted back to an image and exported.