Exam DP-100 topic 2 question 6 discussion

The answer should be "C", the question talks about an environment for model building. Cognitive services is primarily for those with little or no DS skills (Azure Cognitive Services are APIs, SDKs, and services available to help developers build intelligent applications without having direct AI or data science skills or knowledge)

Answer is A: explanation is here: https://docs.microsoft.com/en-us/azure/cognitive-services/big-data/cognitive-services-for-big-data

Azure Cognitive Services Usa modelos ya entrenados. Útil para tareas simples, pero no sirve para entrenar modelos personalizados complejos. B. Azure Data Lake Analytics Es una solución para consultas y análisis de datos a gran escala, no para entrenar modelos. Azure HDInsight con Spark MLlib Bueno para ML tradicional en grandes volúmenes de datos, pero limitado para deep learning y datos no estructurados como video. La respuesta es la D

While Azure Cognitive Services is a powerful suite of APIs and services for adding AI capabilities to applications, it is more focused on providing pre-built models for specific tasks such as vision, speech, language, and decision-making. These services are excellent for integrating AI features into applications quickly and easily, but they may not offer the same level of flexibility and control needed for developing custom deep learning models from scratch. On the other hand, Azure Machine Learning Studio provides a comprehensive environment for developing, training, and deploying custom machine learning models, including deep learning models. It supports various data types and offers tools for data preparation, model training, and deployment, making it a more suitable choice for your needs.

Azure Cognitive Services is for using prebuilt models, not for training

I would pick D and ChapGPT 4 agreed with me. I wrote AI-102 (the updated one) a while back. AI cognitive service is not what you think it is as people discussed here.

Should not it be Azure AI Service?

ANS D - Azure Machine Learning Studio: Azure Machine Learning Studio provides a comprehensive environment for building, training, and deploying machine learning models, including deep learning models. It supports various data types, and you can use popular deep learning frameworks like TensorFlow and PyTorch. This option allows for a more integrated and customizable approach to deep learning model development.

A is correct!

ChatGPT says A.

question talks about the training hence we should go with A

It seems this isnt DP 100 question, but AI 102.

Can we build models with Cognitive services?

i vote for cognitive services, simply because spark is not a better alternative in my opinion. real time processing is not mentioned anywhere

Azure HDInsight with Spark MLib, you can easily build and deploy machine learning models for analysing semi-structured, unstructured, and structured data types. It provides a scalable computing environment that can handle large amounts of data, and you can use various programming languages such as Python, R, and Scala to develop your models. Azure Cognitive Services is a cloud-based platform that provides various APIs and pre-built models for developing applications with natural language processing, computer vision, and speech recognition capabilities. While it can be used to analyse unstructured data such as text and images, it may not be the most suitable choice for analysing semi-structured and structured data such as video recordings and social media logs.

The question references an environment for model building. Cognitive services is primarily for those with little or no DS skills (Azure Cognitive Services are APIs, SDKs, and services available to help developers build intelligent applications without having direct AI or data science skills or knowledge).

option C