How hands-on is this course?

I’m just starting the Supervised ML course, but wanted to ask some of the folks who’ve been doing the course for a while. I was wondering if you actually learn how to build your own machine learning models, because I briefly skimmed the course material and didn’t really find as many hands-on assignments as I hoping to. This was something I know the course was really advertising, so I must have missed something. If you do end up building your own models, could someone let me know what sort of projects they are making/what they’ve seen around? Thanks so much!

They are pretty much guided projects. the main thing is to get the concepts down about how it works, so some code is provided to you. you understand what it does and the theory, but you code smaller pieces. most of the labs build on top of previous problems within the lab to get a final outcome.

they do get longer later in the specialization. for example, you build a image classifier in the neural network course and a mars rover lander in the reinforcement learning course.

I would probably suggest taking this course, maybe taking more courses or learning specific things you want to do, and then doing projects on them.

Hi @Siddharth_Chibber,
Machine Learning contains many concepts and aspects to explore. The course provides fundamental knowledge and guidelines for learners about how different models can be used in various fields. Although the practice labs and assignments look short, they help you recognize the patterns and principles to develop your model in real life.

This is just the backbone. Later on, there are advanced courses that let you learn more about Deep Neural Networks, Image Processing, and Machine Learning in Production. They are also doing short courses about the latest ML technologies like ChatGPT, LangChain, etc. Plus, In every course, there’re additional materials where you can dive deeper into different subjects (research papers, news articles, latest projects, etc…). I also recommend checking out Kaggle if you want to improve your skills after each course.

How hands-on is the course? I would say enough for you to build a basic model, In practice, you have to consider data analysis, deploying, maintaining, and debugging your model in the long run (which requires further training). This is why most complex models you see online require years to be mature.

We all have to start from somewhere, hope it helps.

That makes sense, thanks so much! Are the neural network course and reinforcement learning course both included in the three courses part of the Machine Learning Specialization?

Yeah, the second course is neural networks and basic deep learning. you get a good intro intro to deep learning through this, but a lot more is covered in the deep learning specialization also offered by for example image classification is covered, along with some optimization techniques, but the theory behind each of them are actually covered in the deep learning specialization.

The third and final course goes through unsupervised learning, recommender systems and reinforcement learning. with projects or labs in all three areas.

1 Like