Are you a Data Scientist aspirant? If so, you should be looking at lots of information on the internet to find the best path. Data Science and Artificial Intelligence is a classic combination of computer science, programming and statistics. There are a few certification exams that can be taken to prove your expertise in this field. This blog article is dedicated towards one of the Machine Learning exams (ML) from Microsoft that clearly identifies the appropriate audience as Data Scientists.
This article will help you understand the essential aspects of the exam. This article is also for you if you want to appear and be recognized as one of the few people to pass it very early, like me.
What does the exam actually do?
70-774 is a special exam that focuses on the skills required to create and consume Machine Learning models using Azure Machine Learning Service, Azure Cognitive Services, and public Cloud of Microsoft Azure. This blog article explains the four main objectives and expectations of this exam. The exam includes both case study-based and standalone questions. There could be as many as 40 questions (I got 37). Candidates can complete the exam in just 150 minutes. It’s a typical Microsoft exam. Although you can easily learn all topics of the exam, I recommend that you read these three things before you start:
First: Statistics 101 knowledge is necessary to fully understand the AI algorithms and methods.
Second: Basic Python and R knowledge.
Third: Understanding a few Azure Cloud Services and their functions, not all. But including – Azure Storage Accounts, Azure Data Factorys, Azure HDInsights, Azure Cognitive Services. Azure VM and Azure SQL.
If you have never used Azure before, please register for a free trial account to get started. Cloud Computing with Microsoft Azure Level 2 is a great way to learn Azure. This exam is for Data Scientists. However, it’s for all users. It is a great opportunity to learn important Data Science concepts that will help you build a stronger foundation in Machine Learning technologies.
Objective 1: Prepare data for analysis in Azure Machine Learning and export from Azure Machine Learning
This objective focuses on determining how well a candidate knows about data pre-processing. This is an essential step in any Data Scientist’s journey to create a Machine Learning model that works. If you are new to Data Science, please read the following topics:
Data Sampling
Data Summarization
Univariate vs. Multivariate Datasets in Machine Learning
Impacts of missing/duplicate data and outliers
Methods to deal missing data, duplicate data, and outliers in data
Feature Selection and feature Engineering
All of this requires a lot of effort and can take a lot of time without the right guidance. This course is for those who want to save time and effort. It’s taught by experienced professionals.
You are likely to be familiar with the Azure Machine Learning environment. There are many modules that are used in building a ML model. These modules are preconfigured, packaged pieces of code that perform a specific function. These modules will be questioned in relation to this objective:

It is highly recommended to read the Microsoft Azure documentation and do labs for all modules I have mentioned. This not only applies to this article, but also to the entire objective. It is important to know when and what properties to use a module.