This book gave me the aha moment of understanding that reality IS data, and that’s why algorithms are applicable outside of the domain of computer science.
Great framing. I had read that book, but I hadn’t made that connection (even though I already thought of reality as being pure information—whatever that means).
I feel scared about forgetting the equations that make up the different machine learning algorithms.
In my opinion, this isn’t important—if you can grok the concepts on a gears level, you’re pretty close to having the equations anyways. In real life, no one is stopping you from just refreshing yourself on the equations.
I haven’t taken this course in particular (and Understanding Machine Learning will be among the next three books I review), but I imagine UML would be a good follow-up to firm up the theoretical side. Also, it may be useful to understand how more recent developments work—for example, representing high-dimensional data in low-dimensional latent spaces. One of the most fashionable ways to do this right now is via autoencoders, but you could also get the same effect in other ways.
Edit: apparently downvoting your own comment by mistake doesn’t let you get the full points back after you reupvote, and only lets you get to neutral.
It’s always a bit amazing to me how much I don’t have to remember to be able to work on big software projects. It’s like as long as I know what’s possible, and when it’s applicable, it takes only moments to search for and zero in on specific implementation details.
And yet in this situation, some anxious voice in my head cries, “But do you really know what you’re doing if you can’t remember every detail?!”
So thank you for reassurance on that. Also, thank you for the recommendations!
Great framing. I had read that book, but I hadn’t made that connection (even though I already thought of reality as being pure information—whatever that means).
In my opinion, this isn’t important—if you can grok the concepts on a gears level, you’re pretty close to having the equations anyways. In real life, no one is stopping you from just refreshing yourself on the equations.
I haven’t taken this course in particular (and Understanding Machine Learning will be among the next three books I review), but I imagine UML would be a good follow-up to firm up the theoretical side. Also, it may be useful to understand how more recent developments work—for example, representing high-dimensional data in low-dimensional latent spaces. One of the most fashionable ways to do this right now is via autoencoders, but you could also get the same effect in other ways.
Edit: apparently downvoting your own comment by mistake doesn’t let you get the full points back after you reupvote, and only lets you get to neutral.
It’s always a bit amazing to me how much I don’t have to remember to be able to work on big software projects. It’s like as long as I know what’s possible, and when it’s applicable, it takes only moments to search for and zero in on specific implementation details.
And yet in this situation, some anxious voice in my head cries, “But do you really know what you’re doing if you can’t remember every detail?!”
So thank you for reassurance on that. Also, thank you for the recommendations!