You are designing a solution for a company. The solution will use model training for objective classification. You need to design the solution. What should you recommend?

An Azure Cognitive Services application provides pre-trained AI models via APIs for specific tasks, but it is not typically used for custom model training with unique datasets for objective classification.

A Spark Streaming job is designed for processing live data streams, which is related to data ingestion, transformation, and near real-time analytics, not directly for machine learning model training.

Interactive Spark queries are primarily for ad-hoc data exploration and analysis, not for the systematic process of machine learning model training.

Power BI models are used for data visualization, reporting, and business intelligence, not for machine learning model training.

A Spark application that uses Spark MLlib is an ideal solution for model training for objective classification, as MLlib provides a scalable machine learning library with various classification algorithms suitable for large datasets within the Apache Spark ecosystem.