Cognitive Project Management in AI Free Sample Questions

The CPMAI methodology is iterative. A critical new requirement, such as regulatory compliance for explainability, discovered in a later phase necessitates an iteration. The correct action is to loop back within the current phase or a previous one to address the gap. In this case, returning to Model Development (Phase IV) to build a compliant model is the right approach. Ignoring the requirement (A) or trying to change the project's fundamental goals (B) is inappropriate. Scrapping the model entirely (D) is too drastic; iteration is preferred.

Gradient Boosted Trees, particularly implementations like XGBoost, are well-suited for this problem. They excel at handling complex, non-linear data with high dimensionality and feature interactions, typically yielding high accuracy. Crucially, they also provide built-in feature importance metrics, which directly addresses the sponsor's requirement for insights. Linear Regression assumes linear relationships, which is not the case here. K-Means is an unsupervised clustering algorithm, unsuitable for this supervised prediction task. SVM can handle non-linearity but is less interpretable than tree-based methods.

Apache Kafka is a distributed streaming platform designed to handle high-throughput, real-time data feeds with excellent fault tolerance, making it ideal for ingesting IoT data.

Apache Airflow is a workflow orchestration tool that allows for programmatic authoring, scheduling, and monitoring of complex data pipelines, including both batch and streaming jobs. It satisfies the need to manage complex workflows.

This statement is false. The 'Seven Patterns of AI' (e.g., recognition, prediction, conversation) are high-level problem archetypes. They are most critically applied during Phase I (Business Understanding) to frame the business problem in AI terms and determine the general approach, long before specific algorithm selection occurs in Phase IV.

Recall (or sensitivity) measures the model's ability to find all the relevant cases within a dataset. In this scenario, low recall for the 'attrition' class means the model is failing to identify actual at-risk employees (high false negatives). High accuracy is misleading due to class imbalance (most employees don't leave). The primary business risk is that the company will miss the opportunity to retain valuable talent because the model fails to flag them. Option A describes low precision (high false positives).

The F-beta score is a generalization of the F1-score that allows for weighting precision over recall, or vice versa. A beta value less than 1 (e.g., F0.5-score) gives more weight to precision (minimizing False Positives), while a beta greater than 1 gives more weight to recall (minimizing False Negatives). Since a default (False Positive) is five times more costly, the model should prioritize precision. Therefore, an F-beta score with beta < 1 is the most appropriate metric. The F1-score gives equal weight to both. AUC is a general measure of separability and does not directly account for asymmetric costs.

Data lineage is the process of understanding, recording, and visualizing data as it flows from data sources to consumption. This includes all transformations the data underwent along the way. Implementing robust data lineage is essential for meeting the requirements of traceability, auditing, and debugging AI models. Data masking and anonymization are privacy techniques, and data augmentation is a method for increasing training data size.

A feature store is a central repository for features used in machine learning. Its primary purpose is to solve the problem of feature consistency between training and serving, reduce duplicate feature engineering work across teams, and provide a reliable, low-latency source of feature data for real-time inference. This centralization is a key component of a mature MLOps strategy.

With limited specific data, training an LLM from scratch is infeasible. The most effective strategy is to leverage a powerful pre-trained model and augment its knowledge with domain-specific information. Retrieval-Augmented Generation (RAG) is designed for this; it retrieves relevant information from a knowledge base (like product documentation) and provides it to the LLM as context to generate an accurate answer. This is faster, cheaper, and more effective than fine-tuning with insufficient data or building from scratch.

This is the most accurate description. SHAP (SHapley Additive exPlanations) is a game theory-based approach to explain the output of any machine learning model. It connects optimal credit allocation with local explanations using the classic Shapley values. For a specific prediction, it shows which features pushed the prediction higher or lower and by how much, providing crucial local interpretability for 'black box' models.