An introduction to Time Series Forecasting using a case study of predicting sales of an Ecuadorian store chain

Time series forecasting is a vital task in many industries, as it allows for the prediction of future trends and patterns. This paper represents a concise guide to time series analysis and forecasting, including data preprocessing, feature generation, Exploratory Data Analysis and creating basic models using Exponential Smoothing and SARIMA(X). The models and their hyperparameters are evaluated using a self-built and published Python package. After a theoretical introduction, a case study using a retail store sales dataset from Kaggle is performed. By following the steps outlined in this paper, people concerned with such problems can effectively analyze time series data and make informed decisions for their respective business.