Update README.md

README.md

@@ -45,17 +45,17 @@ attracted the documents to the dense area are the topic words.
 **3. Find dense areas of documents using [HDBSCAN](https://github.com/scikit-learn-contrib/hdbscan).**
 >The colored areas are the dense areas of documents. Red points are outliers that do not belong to a specific cluster.
 
-![HDBSCAN Document Clusters](images/hdbscan_docs.png)
+![HDBSCAN Document Clusters](https://github.com/ddangelov/Top2Vec/blob/master/images/hdbscan_docs.png)
 
 **4. For each dense area calculate the centroid of document vectors in original dimension, this is the topic vector.**
 >The red points are outlier documents and do not get used for calculating the topic vector. The purple points are the document vectors that belong to a dense area, from which the topic vector is calculated. 
 
-![Topic Vector](images/topic_vector.svg)
+![Topic Vector](https://github.com/ddangelov/Top2Vec/blob/master/images/topic_vector.svg)
 
 **5. Find n-closest word vectors to the resulting topic vector**
 >The closest word vectors in order of proximity become the topic words. 
 
-![Topic Words](images/topic_words.svg)
+![Topic Words](https://github.com/ddangelov/Top2Vec/blob/master/images/topic_words.svg)
 
 Installation
 ------------