The problem that will be solved in this written work refers to forecasting sales of distinguished, multinational retail corporation called Walmart, which uses machine learning algorithms in order to improve own business performance. Walmart Recruiting – Store Sales Forecasting data set will be used for the purpose of this work. It has been downloaded from internet platform Kaggle, where public competitions in creating machine learning models take place. The first thing to be presented is theoretical aspect of machine learning throughout process of data mining. In experimental part of work, which refers to realistic case study, explanation of the problem, exploratory analysis and data preparation process will be presented. Afterwards, creation of different models of machine learning is being approached in order to find the best one among them through a comparative analysis of the evaluation results achieved using the appropriate measure. In conclusions, the results of this research will be analysed and explained, additionally giving guidelines for using models in realistic environment. Experimental evaluation will be implemented in programming language Python.