As Industry 4.0 progresses, the industry is undergoing a digital transformation that offers a multitude of new opportunities through the emergence of new technologies. One enabler technology of Industry 4.0 is machine learning (ML), which has become increasingly established as a subfield of artificial intelligence (AI) in recent years. Due to its diverse applicability for solving complex tasks and problems, ML is used for predicting and calculating values, recognizing correlations and optimization purposes, which promises technical and economic advantages. To support decision-making, the supervised ML algorithm Support Vector Machines (SVM) can classify data with respect to more diverse characteristics and features. This master thesis has the goal of solving a classification problem of time series data from sensors used in the mining industry by developing a suitable SVM to reliably classify data and derive from in the corresponding work steps using SVM. In this framework, statistical data analysis, data pre-processing and determination of training and test data were performed. Furthermore, the performances of different kernel functions for the representation of ''Cross Validation Accuracies'' with different kernel method were evaluated and the optimization of the results was performed