Review of google AI education: free machine learning course from google
Google is a company that it's easy to assume most of you know about. The fact is that you probably use or have used at least one of their many services. But it was only a month ago that I discovered one of its important services provided to the community: oh google education, and today we'll talk about it. Today we will review google AI education and the course I took based on my opinion
To do an honest review of this newly discovered service, I needed to explore in-depth the courses provided by Google on the subject last month. However, as there are many courses and classes, I decided to focus on a specific course, with the name:
Machine Learning Crash CourseGoogle AI education
google ai education is where the "Machine Learning Crash Course with Tensorflow" course is located, but before talking about the specific course that I took as impartially as possible, I need to give the attention that google AI education deserves, as it I was surprised at how much content and organization they put into the system. As a content creator and programmer, I respect the effort and affection the company has put into the site, even if some of the content is made to sell their other products (more on that later), the quantity and quality of free content is impressive, and with It sure adds a lot of value to people who use it, even if they don't use other products that it recommends during their courses.
One of the factors mentioned deserves special attention, the organization. Google AI education has the most diverse content and it categorizes them for the most diverse levels, groups, and learning objectives.
When I talk about organization I want to talk about the filters that google AI education puts to organize your content, going far beyond a filter with alphabetical order. The filter separates what you are into:
- Business decision maker: Someone who probably wants to know if their business needs AI or not, what benefits it would have, and so on. With 13 items in this category. Including podcasts, guides, interactive content, courses, and glossaries;
- Curious cat: Someone who is curious about AI and would like to know more about it. With 20 items in this category. Including guides, concept overview, podcast, courses, interactive content, hands-on, competition, and glossaries;
- Data Scientist: The professional in the field of data science. With 29 items on the subject. Including guides, concept overview, podcast, courses, interactive content, hands-on, competition, glossaries, concept overview, and documentation;
- Researcher: Specialized professional who wants to know more about the area, with more substantial and/or academic information. With 19 items in this category. Including guides, concept overview, courses, hands-on, glossaries, concept overview, and documentation and samples;
- Software Engineer: Software engineering professional. With 36 items in this category. Including podcasts, guides, interactive content, courses, glossaries, concept overview, videos, hands-on, etc.
- Student: Focused on students, mainly focused on research on the subject and content for higher education students with interactive courses. With 34 items in this category. Including guides, concept overview, podcast, courses, interactive content, hands-on, competition, glossaries, concept overview, and documentation;
All of the above items can be combined making interesting changes to the content presented. What I mean by this is that you don't have to choose between Curious cat and student if, in doubt, the filter allows you to select both.
And that's not all. The filter also separates content type
- Ex: (13-item courses, 12-item documentation, 1-item chia, 10-item interactive content, 9-item code examples, and 18-item tutorials and code labs, and 9-item videos.
And the filter can be separated in the development stage.
- E.g. (Data collection with 15 items, data preparation with 19 items, Idea development with 23 items, model construction with 32 items, model development with 25 items, and model evaluation with 29 items). Remembering that all elements can be combined with each other for better learning and organization. Also, so far all the content I've found on the site is free.
Machine Learning Crash Course with TensorFlow APIsMachine
Learning Crash Course with TensorFlow APIs was one of several courses on Google AI education and was the course I took. This course promises to feature real-world case studies, hands-on exercises, and video lessons to help you learn Machine Learning yourself, and in my experience, it delivers on what it promises. However, some caveats must be made.
First, it is necessary to pay attention to the prerequisites. Link here.
Some prior mathematical knowledge is required to take the course and introductory classes on machine learning. Knowledge such as Algebra, trigonometry, statistics, Python programming, and terminal usage are some of the topics you will find in the aforementioned link. However, it is not necessary to be discouraged, the good news is that on the link itself there are reference materials to forward you if you do not know
some of these prerequisites. But it is important to note them, as it is possible that you will have difficulty later in the course if you have not gone through this process.
Another caveat is as follows:
The course teaches a lot of theory, which is important, in a generalized way. Which implies being able to apply it in various ways. However, in the practical part, Tensor Flow is used, as the name of the course says. Tensor Flow is an open-source platform owned by Google. Google provides a lot of free content, and to take the course even in the practical parts that use Tensor Flow you don't need to pay anything. But if you consider google's product offering a disadvantage, you should be aware. Despite this, I recommend taking the course, as the theoretical part adds a lot of value.
I personally don't find the use of Tensor Flow a problem, as a company for sure google recognizes its product as the best, in addition to knowing it better than anyone else and needs to earn something in the end for the sea of original quality content provided, however, I thought it was important to mention this in case it could be important to someone.
The course itself impressed me. It's the first time I take a google course, and it inspired me to do this review. I felt that when I finished the course I had grown a little as a professional, despite knowing that I have a long way to go.
I don't think I know everything about machine learning, but I sure do know more than I did before taking the course. Also, as Google provides other complementary courses, I know I can learn even more.
Now I'm going to pay more attention to the organization within the course and describe it, as I think it's important for learning.
As you start each module, you are presented with learning objectives. It works as a "preview" of what will be learned in the class and they receive points for that, in addition to the modules having a time estimate for each of them. Which comes in handy for anyone.
At the end of each class, scrolling to the bottom of the page, you are presented with the keywords of the class. These are important terms that help in the review. So, if you have any questions about a term, you can easily let them know, each term has a dedicated page with explanations.
The overall organization is good, with only one complaint to make. The complaint is that: Even logged into your Google account, the system does not save the classes you have already seen and if you close and open the tab again, you will not return from the class you stopped. I consider this a problem because it relies on people's memories, or the person leaves a tab open all the time. In my opinion, this could be optimized within the system itself.
the course description says it has 15 hours of content, but in my experience, the course is three to five hours less. Despite this, it has a lot of content with experts on the subject and more than 30 activities.
The course is divided into 25 modules from introduction to the next steps. It goes through data collection and classification, generalization, representations, some technical elements, formulas, real cases, practical tests, documentation programming, among other things, until completion. In my opinion, it is a great complete course of quality and free.
In summary, the course is very good, with quality content and great organization.