armanboyaci

Should we validate before we verify the software?

What happens if you have a prior knowledge for $k$ as a probability distribution?

Being able to apply statistics is like having a secret superpower.

I totally with this sentence. BUT If you ask for my opinion, merely knowing a list of statistical formulas is not very helpful. Most of the time, people don’t remember the underlying assumptions, so there is a fair chance they will use them in inappropriate situations.

I recommend watching these two YouTube videos. The presenters advocate using simulation/bootstrapping/shuffling methods instead of memorizing formulas.

Jake Vanderplas - Statistics for Hackers https://www.youtube.com/watch?v=Iq9DzN6mvYA

John Rauser - Statistics Without the Agonizing Pain https://www.youtube.com/watch?v=5Dnw46eC-0o

I believe your description of software development is highly aligned with the ideas of peter naur, programming as theory building.

https://pages.cs.wisc.edu/~remzi/Naur.pdf

I wish there was two types of ipynb files, one for file with just code and markdown (for example ipynbc), and one for keeping code+markdown+results.

I believe you can achieve that if you use jupytext library, right?

I am keeping my CV on overleaf and it is very convenient. Did you try that option?

https://www.amazon.com/Solution-Selling-Fieldbook-Practical-...

If you are planning to build enterprise software solutions then I highly recommend this book. It contains very helpfull checklists and templates.

I recommend this blog: https://yetanothermathprogrammingconsultant.blogspot.com/?m=...

Not all posts are business related but you can learn many practical tricks hard to find in books.

You can compute the max-flow of an undirected graph. The edges have capacities and in the undirected case you assume that capacity can be used in both 'directions'.

I found this explanation: https://www.promptingguide.ai/techniques/rag

General-purpose language models can be fine-tuned to achieve several common tasks such as sentiment analysis and named entity recognition. These tasks generally don't require additional background knowledge.

For more complex and knowledge-intensive tasks, it's possible to build a language model-based system that accesses external knowledge sources to complete tasks. This enables more factual consistency, improves reliability of the generated responses, and helps to mitigate the problem of "hallucination".

Meta AI researchers introduced a method called Retrieval Augmented Generation (RAG) to address such knowledge-intensive tasks. RAG combines an information retrieval component with a text generator model. RAG can be fine-tuned and its internal knowledge can be modified in an efficient manner and without needing retraining of the entire model.

Very good and important observation. In his talk "Simple made easy" [1] Rich Hickey defines simple as opposite of complex and easy as opposite of hard.

The easiness is relative (as you described) and depends on the things you are familiar with. For example, Docker containers and k8s stuff is easy (for you), and GraphQL is hard (for you).

The simplicity should be assessed (somehow) more objectively.

[1] https://www.youtube.com/watch?v=SxdOUGdseq4

I am not sure but another reason could be the mortgage duration. In US 30-year is very common, according to my quick google search in Canada most mortgages are limited to a 25-year amortization period.

You are right.

https://news.ycombinator.com/item?id=34643811

Reminded me this blog post about the complexity cost of trying to cover all the edge cases, using the example of recipe writing!

https://buttondown.email/hillelwayne/archive/edge-case-poiso...

I am seeing multi cars in the screenshots, maybe those are controlled by the players :)

I am surprised when I realized that the author is the same professor in Georgia Tech who wrote this wonderful free book about warehouse planning: https://www.warehouse-science.com/

Unfortunetely he passed away in 2019.

I really like the content then I realized that the lecturer is an ex-CEO of Wealthfront (one of the earliest Robo-advisory company). That totally makes sense considering the practical advices (instead of theoretical discussions) in the course.

Super interesting. Do you know how much advantage a player using this information gets against a 'naive' player non-using it?

Is your work publicly available? If you are interested, I've tweeted my notes on 'Guess Who' game: [https://twitter.com/arman_boyaci/status/1463139888489025537?...]

If you are interested in graph classes, there is another (more mature?) resource: https://www.graphclasses.org/

I am distracted with the example provided in the post. I am pretty sure that most pandas users will just put course_ids on the columns. I mean the shape of the user_courses dataframe is not suitable for the task.

  (user_courses
     .set_index(["student_id", 
                 "course_id"])
     .unstack()
     .apply(lambda x: x+1))

Other learning paradigms are about minimization; reinforcement learning is about maximization.

I don't see why this is important.

I have a friend who also tracks his life: https://www.denizcemonduygu.com/portfolio/vital-signs/

What you see is some of the daily data I’ve collected during the last 7 years of my life (2,498 days between 01/03/2014 and 01/01/2021) charted as basically as possible in order to allow global readings and comparisons. (A nice method is just to scroll up and down with the mouse cursor pointing to the time of interest.) I update it every January. I started to log some topics later on, and there is a period where I did not log mood for some technical reasons in mid-2015. I have a lot more data on medical issues, food, people, and daily activities, which I chose not to show here.

For example, when X plays a corner, there is a possibility that your heuristic may play another corner right? In that situation X wins.

Actually, the first moves of the players determines the result. And because of the symetries there are just 12 first moves.

My tweets about this topic: https://twitter.com/arman_boyaci/status/1442436923960827913?...

Currently I am a happy excalidraw user. They look very similar. For which uses cases tldraw is better suited?

This article might help: "Is a Dataframe Just a Table?" https://drops.dagstuhl.de/opus/volltexte/2020/11960/pdf/OASI...

The content of your website reminded me this service: https://linkkle.com/ demo page: https://linkkle.com/paul

Some similar "AI" challenges: - https://www.pommerman.com/ - https://play.battlesnake.com/ - http://fruitbots.org/ - https://nethackchallenge.com/

Did you mean the game "Hexxagon"?