This talk explores various machine learning approaches to map bark beetle infested forests in Croatia, a threat to forest ecosystems. The study utilizes open-source software like QGIS and SAGA GIS, employing Copernicus data from the Sentinel 2 satellite imagery. Machine learning methods investigated include maximum likelihood classification, minimum distance, decision tree, K Nearest Neighbor, random forest, support vector machine, spectral angle mapper, and Normal Bayes. Among these, maximum likelihood classification is regarded as highly accurate and is commonly used for classifying remotely sensed data. Minimum distance classification is a simple template matching technique, while decision trees are used to identify strategies for achieving specific goals, making them valuable in machine learning. K Nearest Neighbor classifies data points based on the similarity of their neighbors. Random forests, on the other hand, construct multiple decision trees for tasks like classification and regression. Support vector machines are robust prediction tools for classification and regression. The study also explores spectral angle mapper, which measures spectral similarity, and Bayesian networks, specifically Normal Bayes, which uses probabilistic graphical models for predictions. The research evaluates each method using error matrices and compares their performance. The error matrices include a Kappa value, a statistic used to measure inter-rater reliability for qualitative items. All analyses are conducted in the Primorsko-goranska county of the Republic of Croatia. In summary, this paper investigates various machine learning methods for mapping bark beetle infested forests in Croatia using Sentinel 2 satellite imagery and open-source software. The study compares the performance of these methods through error matrices and Kappa values, aiming to find the most accurate approach for addressing this ecological threat in a specific geographic region.