About

This repository represents an academic workshop of data mining course. It contains a practical assignment to get in depth with both supervised and unsupervised learning.

Supervised learning :

The objectives learnt are :

• Visualizing the dataset
• Using naive bayes model and learning its prinicples
• Implementing a method that splits dataset into training and test datasets ( A manual implementation of sklearn train_test_split function )
• Training the model using different training dataset size
• Calculating errors and scores in each case
• Cross validation
• Using Random Forest model

You can find the notebook here : https://github.com/BenrhayemRacem/GL4_TP_DATA_MINING/tree/supervised_learning

Unsupervised learning :

The objectives learnt are :

• Visualizing the dataset
• Using kmeans model and learning its prinicples
• Calculating the silhouette score
• Drawing the dendrogram with hierarchical agglomerative clustering algorithm (HAC)
• Using the Principal Component Analysis (PCA)
• Using an Agglomerative Clustering (AGNES) and drawing its dendrogram
• Comparing HAC and Agglomerative Clustering results with the kmeans using crosstab
• Implementing a manual DIANA ( DIvisie ANAlysis) approach based on kmeans

You can find the notebook here : https://github.com/BenrhayemRacem/GL4_TP_DATA_MINING/tree/unsupervised_learning