Deep Reinforcement Learning

The content is great although a bit dated on some topic, and sometimes the videos seems to come from other modules and refer to content not seen in the nano-degree, however overall the experience is good and the exercise helps a lot.

The content is great although a bit dated on some topic, and sometimes the videos seems to come from other modules and refer to content not seen in the nano-degree, however overall the experience is good and the exercise helps a lot.

Great course to understand modern algorithms in RL.

Great course to understand modern algorithms in RL.

Considering that to buy this course you need to pay Udacity subscription at least one month (250$), this course is simply a scam: Pytorch codes from 6 years ago! That is, the course is still exactly the same from the day it was released, and today (as of 2024) it is outdated. What's more, in some videos they name certain Git hub repositories that have changed. This appears in the actor-critic methods section, which is literally a man marketing himself and reading the code without explaining it, as if he were reading poems. Generally poor explanations of videos of less than 5 minutes, unless you consult the papers and the book that offers, if you start watching the videos you learn rather little. Very poor explanations of programming codes. Unpractical exercises. You have to solve DL problems from already built environments, that is, the practical part is limited to solve video game type environments that have no real life utility. Anywhere is there an example of how to build your own environment? → NO, Impossible deliverables. If you give me a few GPU hours and I have to build my own models and test infinite hyperparameters how am I going to perform that task, if all I can do is copy a code from the internet that I know works so I don't spend those hours? They give very few GPU hours to solve the problems, and these have a lot to do with launching a model, waiting, changing hyperparameters etc. The only good thing about the course is the papers and github repos it redirects to, which are useful. I mean, you pay an absurd amount of money for some internet links. My first and last course at Udacity :)

Considering that to buy this course you need to pay Udacity subscription at least one month (250$), this course is simply a scam: Pytorch codes from 6 years ago! That is, the course is still exactly the same from the day it was released, and today (as of 2024) it is outdated. What's more, in some videos they name certain Git hub repositories that have changed. This appears in the actor-critic methods section, which is literally a man marketing himself and reading the code without explaining it, as if he were reading poems. Generally poor explanations of videos of less than 5 minutes, unless you consult the papers and the book that offers, if you start watching the videos you learn rather little. Very poor explanations of programming codes. Unpractical exercises. You have to solve DL problems from already built environments, that is, the practical part is limited to solve video game type environments that have no real life utility. Anywhere is there an example of how to build your own environment? → NO, Impossible deliverables. If you give me a few GPU hours and I have to build my own models and test infinite hyperparameters how am I going to perform that task, if all I can do is copy a code from the internet that I know works so I don't spend those hours? They give very few GPU hours to solve the problems, and these have a lot to do with launching a model, waiting, changing hyperparameters etc. The only good thing about the course is the papers and github repos it redirects to, which are useful. I mean, you pay an absurd amount of money for some internet links. My first and last course at Udacity :)

Amazing course with the most advanced knowledge about deep RL, yet presented in so pleasant way, that it was such a great adventure to learn and gain new skills with Deep Reinforcement Learning Course! Thank you very much.

Amazing course with the most advanced knowledge about deep RL, yet presented in so pleasant way, that it was such a great adventure to learn and gain new skills with Deep Reinforcement Learning Course! Thank you very much.

The course explains even the heavy concepts very easily.

The course explains even the heavy concepts very easily.