Optimize marketing strategy to launch a targeted campaign by performing customer segmentation
Developed clustering models to perform market segmentation and identify customer needs for a targeted marketing campaign.
Performed exploratory data analysis and visualized customers dataset using distplot (combination between histogram and Kernel Density Estimate) and heatmap correlation using Seaborn and Matplotlib to understand the customer behaviors and their features.
Engineered features using Pandas and Scikit-Learn (data cleaning and standardization) to feed it into machine learning models.
Optimized Kmeans and Principle Component Analysis (PCA) with the elbow method to find the optimal number of clusters using Scikit-Learn for market segmentation.
Built auto encoder using TensorFlow to do the dimensionality reduction; it decreased the original 17 features into 10, then used PCA to drop it down to 2, resulting in 4 different groups of customers.