In order to apply methods and criteria, you must follow some steps. Firstly, define your problem and objective; what are you trying to achieve with data analytics? What is the type and complexity of the issue? Additionally, you should consider the features and variables of your data, as well as the assumptions and limitations of your models or algorithms. Secondly, choose models or algorithms that are suitable for your data and task - this could be existing models from libraries or frameworks, or custom ones. You can also experiment with different variations or combinations, such as ensemble methods or hyperparameter tuning. Thirdly, select data for training and testing - this could be all your data, a sample or subset of it, synthetic or generated data, or augmented data with additional features or transformations. Additionally, split the data into different proportions or folds depending on the cross-validation method you want to use. Fourthly, choose metrics and standards based on your models or algorithms; one or more of the methods and criteria mentioned above can be used, as well as existing ones from literature or industry. Finally, run your models or algorithms and compare their results - this can be done manually with tools or software that automate the process. You can also visualize or summarize results in tables, charts, or reports.