Microsoft Azure AI Fundamentals (AI-900) Practice Exam

To predict the arrival time of a flight based on snowfall at an airport, which modeling approach is appropriate?

1. Classification

2. Time series analysis

3. Regression

4. Descriptive analysis

The correct answer is: Regression

In this scenario, regression is the appropriate modeling approach because the goal is to predict a continuous variable, which is the arrival time of a flight. Regression techniques are specifically designed to handle problems where the outcome is numerical and not limited to categories. In this case, the arrival time is influenced by snowfall, which is a quantitative variable. Through regression analysis, one can establish a relationship between snowfall data and the predicted flight arrival time. This allows the model to output a specific time value as a prediction, rather than categorizing the outcome or making time-based predictions which are characteristic of other methods. While classification is used for problems where the output is a category or label, such as identifying if a flight is on time or late, it would not apply to predicting exact times. Time series analysis could be relevant if the data were solely based on time-based trends over a series of observations, but since the focus here is on a direct relationship between snowfall and arrival time, regression is more suitable. Descriptive analysis, on the other hand, involves summarizing existing data rather than forecasting or predicting future outcomes, making it inappropriate for this predictive task.