Fast Editorial Roundup – What to Expect, and When
Timetable for AI and Deep Learning Study Resources:
This week, I’ve been working on the simplest, smallest-possible useful release for you; the Seven Essential Equations Cribsheet. It’s nearly there, but not quite.
This means you’ll get an email soon with a link to click – and another Opt-In form – to access the Cribsheet.
Editorial Calendar (Draft):
Here’s the short-term calendar:
- This Week (Aug. 25 – 26): The Cribsheet for The Seven Essential Machine Learning Equations,
- Next Week (Aug. 27 – 31): Labor Day Weekend Immersive Study Guide: What to Read and Watch, Prioritized List,
- Starting After Labor Day (Sept. 7th and onward): School’s in Session: Fundamentals on Statistical Mechanics for Machine Learning (a blog series),
- St. Crispin’s Day (Oct. 25): Opportunities – from getting earliest drafts of my new book to a private “Inner Circle” study group, and
- Christmas Vacation (Dec. 26 – Jan. 9): Private Immersive. (You weren’t really going to just have fun over the holidays, were you?)
Next substantial blogpost: How to Shorten the Runway, responding to a question from one of you.
Earlier this week, R.S. commented on the Long Runway notion:
One challenge I face is this – I feel in the workplace today, one is expected to have a short runway – always – to be competitive, to deliver value to the internal & external clients. I would love to have a long runway, but job security is a question on my mind.
Yup, that’s the crux of the matter: how can we shorten the runway?
That’s why I’m hustling to get the Cribsheet ready for you as fast as you can. The Seven Essential Equations is the map. The Cribsheet is the legend for reading the map.
Just having a map in hand doesn’t get you from the west of the Rockies (Idaho, Montana, and Wyoming) to the machine learning gold rush in California. But it’s a whole lot better to have a map than to not have one at all.
The next thing I’m going to get to you will be a prioritized reading list. Just ASAP.
In between these two things … I’m feeling a rant coming on. You’ll hear from me on it. You definitely will.
Help Me Help You – Fast Reply, Please!
I’ve got two versions of the Cribsheet going on.
- Version 1 – Bare-Bones: This is JUST “how to read the equations,” with little tiny bit of description about what the equation is, what it does, where it fits into the grand scheme of things (and why you should know it, of course), and
- Version 2 – Extended-Play: All the above PLUS:
- Examples: Examples – going from completely abstract equation-thinking to something sort of tangible, and
- Diagrams: I so totally believe in “See Spot run.” With pictures.
You folks are bright, God knows; you’re hugely intelligent. Focused. Hard-working.
You’re also very short on both sleep and time.
The more I can go from abstract-esoteric-equation-land to a diagram that tells the story, the better it is for ALL of us. (Me, as well as you. I get tremendous clarity on building these things.)
But … they take time.
So, do you want the Version 1 Cribsheet ASAP, or are you willing to wait another week to get Version 2?
Do you want Version 1 AND Version 2?
Do examples and diagrams really help you? I think they do, but until you give me feedback, I don’t know for sure.
And if I’m hustling to produce Version 2 of the Cribsheet, that means a delay on the Prioritized Reading / Watching List.
So what are YOUR priorities? Which do you want first, the Reading / Watching List or Version 2?
Comment, please!
And a big, fat THANK YOU!
Live free or die, my friend –
AJ Maren
Live free or die: Death is not the worst of evils.
Attr. to Gen. John Stark, American Revolutionary War
4 thoughts on “Fast Editorial Roundup – What to Expect, and When”
I’d rather wait another week for Version 2. Thanks so much for doing this!
Thanks, Luis, for being first vote in, much appreciated! With you and Gregg B. (see following comment) both voting for the “extended play” version of the Cribsheet (“legend” to the Seven Equations “map”), that’s what we’ll do. So, I’ll be nose-down on getting this done over the Labor Day weekend!
Point:
2. Next Week (Aug. 27 – 31): Labor Day Weekend Immersive Study Guide: What to Read and Watch, Prioritized List,
Response:
I found the book by Hastie and Efron, Computer Age Statistical Inference: Algorithms, Evidence, and Data Science to be a must read.
https://web.stanford.edu/~hastie/CASI/
Point:
4. St. Crispin’s Day (Oct. 25): Opportunities – from getting earliest drafts of my new book to a private “Inner Circle” study group, and
Response:
Please do share this, I look forward to reading it and providing feedback.
Point:
So, do you want the Version 1 Cribsheet ASAP, or are you willing to wait another week to get Version 2?
Response:
Version 2
Point:
Do examples and diagrams really help you? I think they do, but until you give me feedback, I don’t know for sure.
Examples and diagrams help build intuition which is critical in building understanding. That is why I am an advocate of Sal Khan and his work at Khan Academy, where he places alot of effort in building an intuitive understanding of what is being done through extensive graphical illustrations. I would recommend doing similar tutorials – in fact you can use the tools Sal uses and possibly incorporate them into the Khan Academy material.
Hi, Gregg – thank you for the detailed and valuable response!
You’re the second person (along with Luis, see above) to vote for the “Extended Play” version of the Cribsheet (the “legend” for the Seven Equations “map”), and so that’s what I’ll do. Nobody is saying that they need the short version overnight, so … this coming weekend, I’ll be getting that done.
Once I get the Cribsheet out, I’d so appreciate your inputs on the level of diagrams / figures / explanations / examples. That won’t be for another week, though.
I’ve just now looked at the book that you’ve recommended, “Computer Ages Statistical Inference,” and it seems really good! (Here’s the Amazon link, with the Look-Inside option: https://www.amazon.com/Computer-Age-Statistical-Inference-Mathematical/dp/1107149894/ref=sr_1_1?s=books&ie=UTF8&qid=1504116415&sr=1-1&keywords=computer+age+statistical+inference#reader_1107149894).
Efron and Hastie’s book would be a great complement to what I’m working on; a very different perspective. They address some of the topics that we’ll be going through; some examples:
1) Expectation Maximization – they address in Chpt. 9.5; I expect to give more attention to this method,
2) All of Chapt. 13 is on the Objective Bayes Inference and MCMC (Markov chain Monte Carlo), with a little bit of Gibbs sampling; and
3) Chapt. 18 deals with neural networks and deep learning.
All in all, the topics that we’ll address here comprise about 20% of what they cover, and I feel that this is a good complement. We’ll put most of our emphasis on neural networks, deep learning, and more attention leading up to variational inference.
However, I looked a bit at their Chapt. 13 (Bayesian inference), and their writing style is very elegant and lucid, easy to follow and with great insights. I’m certainly going to read that chapter in depth, and several of their other chapters as well.
Thanks for the great recommendation!