Help me out

Course Q&A Machine Learning Specialization MLS Resources
1

i am taking the machine learning specialization course and i am understanding the theory of things i just want advice how do i apply these concepts in code on my own into projects or anything to practice i am aware of kaggle but everything seems daunting how do i start

2

You should complete the entire Machine Learning Specialization first, then maybe take one of the “deployment” courses if you’re not comfortable with Kaggle.

3

Hi @Aman_Mehra,

I googled “step by step ML project walkthrough” and found this walkthrough. Here are my suggestions:

  1. Give it one week to do these:

    1. Remember that, you are learning, so no need to think big yet. You are learning skills useful for your whole life, so no rush. You are training yourself, so repetitive work is good, and ultra care to details is wonderful.

    2. Follow the walkthrough, understand every single sentence, and repeat every single step yourself in your own jupyter notebook.

    3. Understand every line of codes by googling. For example,
      image
      Google the function name “pd.read_csv” for its official documentation page, spend <10 minutes to read as much as you can (if not finished, come back next time if you don’t understand the function again), and experiment it in anyway possible, such as reading another csv of your own.
      Actually making changes and doing something help you remember, no matter how obvious or dumb it seems. In fact, it is okay to do dumb things because only you know it, so don’t just read and feel OK, but do something.
      If official doc is not good enough for you, go back to google, and check out some stackoverflow results. Official doc is general, stackoverflow can be more case-specific. We need both - we rely on official doc for the general, and we can get some specific from stackoverflow and our own work.

    4. Understand every concept by (a) googling more discussions with names; AND (b) referring back to lectures or your notes; AND (c) playing with the codes that implement or show the concepts; AND (d) any other way you prefer - do everything you can think of and don’t feel satisfied before done doing everything

    5. Make notes to remind yourself what steps you need to take in your next project - don’t put all burdens on your brain, but most is OK.

    6. Remind yourself again - dive in as deep as you can hold, and experience as much as you can afford. Don’t skip steps, and don’t just take shortcuts (but do try). Both success and failure are important for us, because we need to know how to face with both in the real world.

    7. When tired, take a rest, and come back :wink:

  2. Google 5 more walkthroughs, and repeat step 1 for each of them.

  3. Try to start an entry-level kaggle with the codes and concepts you have learnt, with the help of the notes you have made.

Many resources are available on the internet, we only need to reach out, and to practice on how to reach out effectively. :wink: If some googling keywords don’t work out, try more :wink:

Good luck!
Raymond