Regression from Classification Tasks

Requirements Clarification & Assessment

1. Understanding the Problem Types

   • Regression: Predicts a continuous quantity, such as predicting the temperature, stock prices, or real estate values.
   • Classification: Predicts a discrete label, such as determining if an email is spam, or classifying images of animals.

2. Identifying Use Cases for Transformation

   • Determine scenarios where converting a regression task to a classification task is beneficial, like simplifying the problem or aligning with business objectives.

3. Data Characteristics

   • Evaluate the nature of your data, including:
     • Distribution of the continuous variable.
     • Potential thresholds or boundaries for classification.
   • Understand the implications of converting continuous data into discrete categories.

4. Outcome Requirements

   • Define what success looks like for both regression and classification tasks in terms of accuracy, interpretability, and utility.
   • Consider the impact of model choice on the interpretability and usability of results.