Azure AI-900 Exam Dumps|Certification Preparation-Datacloudy

 

In this blog we are going to see about some important points and thumb rules that must be known to clear the AI-900 Exam. Please follow this blog and I am sure you will crack the exam.

Here we are not only going to give the topics but we are going to give a crisp and clear explanation for each topic so that you never forget the concept. And you can also use this blog as 15 minutes refresher after your preparation and before going to the certification exam.

And I am sure, this going to be a "cheat sheet"  for your certification exam. Now let us see the important points,

(*)  Types of AI:

    There are 2 types of AI and they are,

            i) Strong artificial intelligence or General AI:

                    It learns like child and we are far away achieving this.

            ii) Narrow AI or Weak AI:

                    It Focus on Specific task .For example,  Self driving and virtual assistance.

(*) Difference between Traditional Program and Machine Learning:

        The key difference between them are,

•     Traditional Program uses Rules
  
  
•      Machine Learning uses Examples.

(*)  Approaches to build AI solution in Azure:

        There are 3 approaches to build AI solution in Azure, they are

            i)  Use -Pre trained models:

                    Example for this is Azure Cognitive Services

            ii)  Build Simple Models:

                    Without need of data scientist and limited/no code experience.
                    Example: Custom vision, Azure Machine Learning (Automated Feature)

            iii)  Build complex models: 

                    Using data scientist, build own ML model from scratch. 
                    Example: Azure Machine Learning.

(*) Azure Cognitive Service:

        In an overview it comprise of ,

                i) Computer vision

                ii) OCR

                iii) Face API

                iv) Form recognition

                v) NLP

        click here to get brief explanation of the above topics and it is very important in the certification point of view.

(*) Decision Services:

        This makes smarter decisions. Please remember this topic, there will be a question on this topic for sure.

        i) Anomaly Detector:

              Used to find the Fraud.

        ii) Content Moderator:

               Detect unwanted content in the text, that is like  profanity issues in website.

(*) Custom Vision: 

        click here to get crisp and clear details of Custom Vision and the inner topics of Custom vison that needs to be focused for AI-900 certification Exam

(*) Understanding Machine Learning Terminology:

      i) Process:

            

         1. Training : The process of creating the Model

         2. Evaluation: Checking is the Model working?

         3. Inference : Using Model to do predictions in Production.

      ii) Dataset : Data used to create, validate and test the Model.

          Dataset Types:

                    

          1. Training Dataset : Dataset used to create Model.

          2. Validation Dataset: Dataset used to validate the model ( and choose the right algorithm)

          3. Testing Dataset : Dataset used to do the final testing before Deployment.

(*) Model Evaluation - Regression Model and Classification Model:

    Click here to know about this topic in detail manner. As this is an important topic, don't skip it.

(*) Terminology:

        Some of the Key terminology you must be known in the Certification Exam Point of View are given below,

        i) Studio : Website for Azure Machine Learning.

        ii) Workspace : Top level resources for Azure Machine Learning.

        iii) Azure Machine Learning Designer : Drag and Drop interface to create Machine Learning Workflows(canvas)

        iv) Inference Cluster : Deployment machines for Azure Machine Learning 

        v) Model registration :  To store and version your models, Organizing and keep track of your trained models.

(*) Responsible AI Principles:

    This is an very important topic for the Certification Exam. Kindly give more attention to this topic.

    1. Fairness:

            Fair to all group of people. For example "gender" equality.

    2. Reliability and Safety: 

            Continuous working of model under high loads and unexpected situation.

    3. Privacy and Security:

            It deals with the protection of information.

    4. Inclusiveness :

            Nobody should be left out. For Example, The usage of model should not omit physically challenge persons.

    5. Transparency :

            It deals with the Explainability of the Model. Like providing the details what the Model is for.

    6. Accountability:

            The Model should meet the Ethical and Legal standards. For Example: Having a Team that can override the decision made by the AI system.

Thus in this blog, we saw the important points that needs to be taken care to crack the Azure AI-900 Certification Exam. Please go through each and every topic, For some of the important topic links we given under respective headings. Hope this is helpful. You can Follow this blog to show your support by clicking the Follow Button in top right pane of this window.

Thank You!!!