I enter as many Kaggle competitions as I can because they offer an opportunity to solve a wide variety of machine learning problems.One such competition question was Coursera’s final project for their course, “How to win a data science competition. This competition provided the entrant with sales statistics for 33 month to enable time series analysis of sales to be conducted on the final month.

The problem description reads as follows:-

In this competition you will work with a challenging time-series dataset consisting of daily sales data, kindly provided by one of the largest Russian software firms — 1C Company. We are asking you to predict total sales for every product and store in the next month.

The competition question was quite challenging and I shall discuss some of the challenges when I review the technique I used to solve this problem.

The first thing I did was to import the libraries in Python that would be necessary to write the initial stages of the program:-

#artificial-intelligence #machine-learning #python #data-science #time-series-analysis