[Education] Udacity Nanodegree: Machine Learning Engineer

The Nanodegree program from Udacity offers some tech certificates, especially for programmers. One of them is called “Machine Learning Engineer” and with the github Student Developer Pack one month is free. So i decided to enroll this course in my semester break and try it out.

Massive Open Online Courses

I already did some MOOCs before, but to be honest i don´t think they are worth much. While i am a big fan of project based learning and I don´t like traditional classroom teaching style, i still think a MOOC is far away from any projects from the “real” university. I did these MOOCs mostly because i wanted to try it out and to get some certificate, which i can show the non-tech hr managers, if necessary (which never happend). I really did all these Quizzes, but i don´t think there was much, i got out of it.

I attended the free courses:

  • Online Games: Literature, New Media, and Narrative
  • Pattern-Oriented Software Architectures: Programming Mobile Services for Android Handheld Systems

When i started my studies in computer science i wanted to make games (as many others), so i attended the first MOOC. I think it took 2 days to complete all tasks and it was very simple. Well i had a course about game programming at the university and was interested in the topic, so i knew the most things before. The second one was more work, but still kind of unsatisfying. The video lectures were always good, but that is nothing new too me, as i learned a lot with youtube in the basic studies.

I think there is also a huge problem with plagiarism, as the most programming tasks are submitted to github and i am sure people just copy them, without anyone notices it.

Anyway i will give it another try with the Machine Learning Engineer nanodegree at udacity.

Skip the basics

One major problem with MOOCs is the different level of knowledge of the students. In Germany the university professors always complain about this difference between states and even schools. So if this is a problem of local scale, how can this work in MOOCs?

The simple answer is , it doesn´t need to work. In the MOOC you can just skip all the stuff you already know (and even in advanced courses this can be a lot). Everyone can have his own speed. So said i skipped the first optional project, called Titanic Survival Exploration(surprise, surprise) and submitted the first milestone Predicting Boston Housing Prices, after 10 hours work.

Fast reviews

Wow, only 2 hours, after i submitted my first milestone, i got a review and i was much more detailed, then i thought. I updated my answers and re-uploaded the project and the second review was also very fast. So nothing to complain about here.

Related Posts

Leave a reply