Microsoft Azure AI Fundamentals (AI-900) Practice Exam

When training a predictive model for taxi fares, what feature is most relevant?

1. The fare amount

2. The trip distance of individual taxi journeys

3. The number of passengers

4. The driver’s rating

The correct answer is: The trip distance of individual taxi journeys

When training a predictive model for taxi fares, the trip distance of individual taxi journeys is the most relevant feature because it directly influences the fare amount charged to passengers. In typical taxi fare structures, the fare is often calculated based on a combination of factors that include a base charge, a charge per mile or kilometer, and potentially time-based charges for waiting. Hence, the trip distance is a key determinant, as longer distances generally correlate with higher fares. The other features, while they may provide some context, do not have as strong a direct correlation to the fare. For instance, the fare amount itself is the target variable that you are trying to predict, so using it as a feature would be redundant. The number of passengers could influence the fare in some cases, but it is typically a smaller factor compared to distance. The driver’s rating may impact customer satisfaction or choice of taxi, but it does not directly affect the fare calculation itself. Thus, trip distance emerges as the most pertinent feature for accurately predicting taxi fares.