The problem definition for Titanic Survival competition is described here at Kaggle.
Knowing from a training set of samples listing passengers who survived or did not survive the Titanic disaster, the goal is to determine if the passengers in the test dataset survived or not.
Classifying and Categorizing. This helped me to understand the implications or correlation of different classes with our solution goal.
Correlating. Determining correlation among features to check which features within the dataset contribute significantly to the solution goal and also to determine correlation among features other than survival for subsequent goals and workflow stages.
Converting. Converting text categorical values to numeric values to facilitate better working of the classification model.
Completing. Filling in the missing feature values in the data to ensure Model algorithms may work best when there are no missing values.
Correcting. Analyzing the given training dataset for errors or possibly innacurate values or outliers within features and trying to correct these values or exclude the samples containing the errors.
Creating. Creating new features based on an existing feature or a set of features, such that the new feature follows the correlation, conversion, completeness goals.
Charting. Plotting visualization plots and charts to obtain a thorough understanding of the data.