Algorithms for Decision Making

Can you recommend any good books or other resources?

You are right that sometimes people use much more sophisticated algorithms than are really required to solve a problem. My experience is that the simplest approach is often the best one in the long term. There are actually deep connections between problems in control theory and the topics in this book (e.g., LQR control), though these different communities often use different notation. The focus of the book is more on higher-level decision making, but the execution of the decisions---such as motor commands to an actuator---are often best done through PID or some other method that can be found in a control theory textbook.

Control theory is indeed, quite cool, and under-discussed in most circles.

Also, unfortunately, classical control theory is primarily good for linear time-invariant dynamics in the frequency space of the Laplace transform. If you can't locally linearize your model and/or need to learn a model, classical control approaches are underdeveloped, and everyone has switched to optimal control and RL.

After the basic PID controller, what would you consider the most useful tools for control?

For control theory from a mathematician, Fleming, long at the Division of Applied Mathematics at Brown University, there is:

Wendell H. Fleming and Raymond W. Rishel, Deterministic and Stochastic Optimal Control, ISBN 0-387-90155-8, Springer-Verlag, Berlin, 1979.

There is more on optimization, especially from the Hahn-Banach theorem, and also Kalman filtering, in

David G. Luenberger, Optimization by Vector Space Methods, John Wiley and Sons, Inc., New York, 1969.

This is neat - specifically because what you point out is actually true. The primary author is a professor and the secondary author is a Phd graduate - both in the Stanford Aeronautics and Astronautic Department [1][2] - in the Dynamics and Control Group.

A brief skim through of the book shows many references to Control Theory texts.

[1] https://mykel.kochenderfer.com

[2] http://timallanwheeler.com/aboutme/cv/Wheeler_CV_201804.pdf

Yeah, that was a really great book. Also has a well-produced narration available on Audible that I listened to. I recommend it.

I love this book. Listened to it 4-5 times. Lots of really practical application

Such a good book. I learned so much about life from the multi-armed bandit problem.

Reading this at the moment! Fantastic book and easy to read

I still think about this book every time I get on Zillow. Great chapter on buying a home.

I recommend that one to my students. I also recommend Brian's new book titled "The Alignment Problem" and it is cited in the book.

Yes! ATLB is fantastic!

I bumped into this accidentally having ever heard of it. It is one of those pleasures of browsing through a library. Lovely book, easy to understand with a broad range of topics and insights. Recommended!

Just started reading this thanks to your recommendation, it looks like a really entertaining book, thanks!

PS for others, it's on libgen...

I link at the source of this anecdote at the end of my essay on I-Ching: https://www.pa-mar.net/Lifestyle/I-Ching.html

Here is the direct link if you don't care about my stuff (it has been featured on HN already)

http://statweb.stanford.edu/~cgates/PERSI/papers/thinking.pd...

*first two authors (MK and TW) wrote Alg4Opt This new book is in the same format.

I highly recommend this book. I am 100% going to buy their next book too.

Julia is an awesome language and the ecosystem around it is getting better quickly.

As a programmer I usually try to implement some of the algorithms in language of my choice and throw them at some toy problems (it's good there is no shortage of interesting toy problems in this field). It was really fun with Artificial Intelligence: Modern Approach book (ex: implementing a SAT solver from scratch and solving Minesweeper by pure logic is very satisfying).

Anki is basically the most efficient method to study facts. I use a gamified version called Memrise.

Get in a group. Work is a good place to do this. Pick out some books, send a mail round saying that you are running a tech book group. Get folks together and read a chapter a week (this is hard in reality - be cool with everyone and sometimes let chapters slip over a few weeks). Every week someone leads the discussion. Every week someone volunteers to do the next weeks. You meet for an hour to discuss.

I typically take a book like this and read it. Then I think about it. I also think about it as I read it. I ask my friends in person and online if they've read and talk with them about it. Sometimes I'll look for online forums which discuss the book. This is how I make the most out of books like these.

My suggestions, not for this book in particular but technical textbooks in general:

- Get very familiar with the table of contents first. Then note and filter out the sections you are less interested in or don't appear to provide the biggest bang for your buck.

- Annotate, annotate, annotate. Active reading is better than passive reading. I use a 2-in-1 chromebook and a passive pen to highlight in the adobe pdf reader.

- Write detailed notes in addition to annotating (i.e. in a separate notebook or markdown file). Cull/convert your notes into Anki/Memrise digital flashcards. Drill on your phone when you are lying in bed.

it's hard for me to learn something if i don't have something I need to do with the knowledge. it's not that I can't sit down and do it, but things won't stick in my brain.

so I would say, try to find something you want to do that requires the information in the textbook. or else just don't bother reading the textbook and go do something else you actually want to do.

I enjoy typing, so I like to simply type out the entire textbook. It's a bit ridiculous to some, but it slows down the pace that I actually learn and engage with the content. A textbook usually takes me a couple months to type, and I'm never in a rush though finishing chapters is satisfying and an easy goal for an evening. I find it similar to as if I were to attend a lecture on the topic.

The way I did this with a stats book was:

  1. Ten pages a day
  2. Do all the exercises, and redo all the ones you get wrong.
  3. Add anki cards for all the key terms, ideas, formulas, etc. The best textbooks include a summary the end of every chapter that you can double check your notes / anki against. Prefer cloze markup to question / answer cards -- more cards per "fact" but longer retention.
  4. Double check your cards against a source like Wikipedia for completeness / accuracy, since sometimes there's an author bias to content with.

This takes about an hour a day for #1 and #2, and an hour a week for #3/4 plus a few minutes a day extra in the Anki review queue. For this book, you're looking at about 2 months start to finish.

Anki can be a supercharger, if you have notes that are clear and concise.

Since you had an Ask HN regarding machine learning resources, and your uncle has written some machine learning papers, have you considered asking him to teach you or is he really busy?

Looks like a different book, "Algorithms for Optimization", by two of the three authors of "Algorithms for Decision Making".

Sure thing! We wanted to get feedback from the broad community so that we can make it as awesome as possible before it goes to print. ;-)

The authors plan to have the book be free, online, forever!

I think Calibre can convert from pdf to kindle format

A print version is planned with MIT Press for release in the fall! PDF will continue to be available for free forever.

If you wanted to do Cloze questions, you can use saveall.ai (a tool that I built) to generate them automatically from fragments of texts. We also plan to work on automatically generating questions from a larger chunk of text, but it is very difficult. Some people have had some success with GPT-3 but it's unreliable

Wild guess is that the main document LaTeX file is called "dm.tex", so the resulting PDF would be called "dm.pdf".

Making the main file named something more verbose would impact you whenever you try to open it, but in an ideal world, you'd rename it once it gets to the point of it not needing many changes.

This is a very important dimension. See Sec. 1.5 of the book. There is also a side reference to a book that discusses this and other societal implications.