Analysis of crimes in Mexico during 2017 with Machine Learning techniques (Cluster Analysis): Comparison Elbow Method and Silhouette Method
Camacho-Perez Enrique, Arroyo-Velázquez Isaac
During the elective course of "Data Science" taught by Prof. Enrique Camacho at the Faculty of Engineering UADY, we performed a task to understand and practice the K-Means algorithm (an unsupervised learning algorithm) and the selection of optimal number of clusters with the help of the Elbow Method. In this repository we go a little deeper than usual in class and compare it against another algorithm for selecting the optimal number of clusters called the Silhouette Method.
The analysis and procedure are documented in the Jupyter Notebook named ClusterAnalysis_CrimesMexico2017_en.ipynb and in the following Medium's article:
- Python +3.7
- The following libraries:
- NumPy:
pip install numpy
- Pandas:
pip install pandas
- Sci-kit Learn:
pip install scikit-learn
- Folium:
pip install folium
- Seaborn:
pip install seaborn
- NumPy:
- Isaac Arroyo:
- e-mail: isaacarroyov@outlook.com
- Enrique Camacho:
- e-mail: enrique.camacho@gmail.com
- GitHub