Predicting e-commerce audience age and gender

Machine Learning Project

Here is intended to predict age and gender of online customers, from the URL history from an online retailer, thus marketing team could use this information to develop segmented campaigns. The complexity of this problem was the data preparation and the size of the database. The dataset given contained words and numbers, that represent the words given by the URL visited by each customer, and the frequency of these words respectively. After data preparation, models are trained and ready to predict. Here I used Naives Bayes, Random Forest and Linear Regression.

Finally, the predictions are exported into a csv file.

Language: Python | Libraries: Numpy, Pandas, Matplotlib, Seaborn, Nltk, Sklearn.