NCA-AIIO Question 75
Select 3You are working under the supervision of a senior team member to analyze data for training an AI model. During this process, you notice that the dataset contains a high number of missing values in several key features. What should you do to handle this situation effectively while ensuring the AI model’s performance is not compromised?
- A
Discuss with the senior team member whether to impute the missing values using statistical methods like mean or median.
- B
Directly remove all rows containing missing values to simplify the dataset.
- C
Analyze the impact of missing values on the model's target variable and consult with the senior team member on the next steps.
- D
Replace missing values with random numbers to maintain dataset size without consulting the team.
- E
Propose using advanced imputation methods, such as predictive modeling, and get feedback from the senior team member.
Show answer and explanation
Correct answers: A, C, E
Explanation
Handling missing values in a dataset is a critical step in AI model development. The correct approach involves consulting with senior team members to ensure that decisions, such as statistical imputation or advanced techniques, align with project requirements and maintain data integrity. Analyzing the impact of missing values on the target variable is also essential for making informed decisions. Actions like removing rows or introducing random values without consultation can lead to suboptimal results or compromised model performance.
- A. Correct.
Discussing the use of statistical imputation methods like mean or median with a senior team member is a valid approach to address missing values in a structured manner.
- B. Incorrect.
Directly removing all rows with missing values can lead to a significant reduction in data size, which might negatively impact the model’s training. This action should only be considered after consulting the senior team member.
- C. Correct.
Analyzing the impact of missing values on the target variable and consulting the senior team member ensures that the decision aligns with the overall goals of the AI model and data integrity.
- D. Incorrect.
Replacing missing values with random numbers without consulting the team is not a recommended practice, as it introduces noise and may lead to biased model results.
- E. Correct.
Proposing advanced imputation techniques, such as predictive modeling, and seeking feedback from the senior team member is a thoughtful approach to ensure high-quality data preparation.