What is the main goal of the evaluation step in the data mining process?

• A. Selecting the best algorithm
• B. Deploying the model
• C. Evaluating whether your chosen data model fits the underlying data set
• D. Collecting more data

Answer: (C)

Evaluating what your model learned in relation to the actual data is the core aim here. After building a model, you need to check whether the patterns it captured actually reflect the real data and whether it can predict new, unseen cases. This means testing on separate data (or using cross-validation) and using appropriate performance metrics to assess generalization, not just how well it fits the training data. If the model performs well on hold-out data and meets the desired criteria, it’s a good candidate for deployment; if not, you’d revisit features, try different modeling approaches, or collect more data. Choosing an algorithm is typically part of the modeling step, deployment moves you to putting the model into use, and collecting more data happens earlier in the process.