This white paper provides an overview of metrics used in multi-class classification tasks. It discusses various performance indicators that help evaluate and compare different classification models or machine learning techniques. The paper highlights the advantages and disadvantages of these metrics and explains their usage during the development of a classification model. Multi-class classification involves predicting one of several possible classes for a given input. The paper reviews several metrics, including Accuracy, Balanced Accuracy, F1-Score, and Cross-Entropy, explaining how they are calculated and their implications for model evaluation. It also discusses the Matthews Correlation Coefficient (MCC) and Cohen's Kappa, which measure the correlation between predicted and actual classifications. Accuracy is a common metric that measures the proportion of correct predictions. However, it can be misleading in imbalanced datasets. Balanced Accuracy adjusts for class imbalance by averaging the accuracy across all classes. F1-Score is a harmonic mean of Precision and Recall, providing a balanced measure of a model's ability to identify positive instances. Cross-Entropy measures the difference between the predicted and true probability distributions. It is useful for evaluating the quality of predictions in multi-class settings. MCC and Cohen's Kappa are more sophisticated metrics that account for the correlation between predicted and actual classifications, providing a more nuanced evaluation of model performance. The paper concludes that the choice of metric depends on the specific goals of the classification task, with some metrics being more suitable for balanced datasets and others for imbalanced ones. It emphasizes the importance of selecting the appropriate metric to accurately assess the performance of a classification model.This white paper provides an overview of metrics used in multi-class classification tasks. It discusses various performance indicators that help evaluate and compare different classification models or machine learning techniques. The paper highlights the advantages and disadvantages of these metrics and explains their usage during the development of a classification model. Multi-class classification involves predicting one of several possible classes for a given input. The paper reviews several metrics, including Accuracy, Balanced Accuracy, F1-Score, and Cross-Entropy, explaining how they are calculated and their implications for model evaluation. It also discusses the Matthews Correlation Coefficient (MCC) and Cohen's Kappa, which measure the correlation between predicted and actual classifications. Accuracy is a common metric that measures the proportion of correct predictions. However, it can be misleading in imbalanced datasets. Balanced Accuracy adjusts for class imbalance by averaging the accuracy across all classes. F1-Score is a harmonic mean of Precision and Recall, providing a balanced measure of a model's ability to identify positive instances. Cross-Entropy measures the difference between the predicted and true probability distributions. It is useful for evaluating the quality of predictions in multi-class settings. MCC and Cohen's Kappa are more sophisticated metrics that account for the correlation between predicted and actual classifications, providing a more nuanced evaluation of model performance. The paper concludes that the choice of metric depends on the specific goals of the classification task, with some metrics being more suitable for balanced datasets and others for imbalanced ones. It emphasizes the importance of selecting the appropriate metric to accurately assess the performance of a classification model.