sonodewil

Habt ihr euch jemals gedacht "jetzt ne Gute Nacht Geschichte vorgelesen von Manu, das wärs"? ... ich auch nicht, aber hab's trotzdem gemacht, wenn ihr Bock habt: youtu.be/5sPh6vF6Wio

HAPPY EASTER! He is Risen!

DOMINGO DE RESURRECCIÓN

I think we all need some good news with what else is going on in the world. Here is some: as of today, you can pre-order my new book harriman-house.com/authors/robert…

and of course keep in mind the position we applied for REQUIRES hard knowledge in finance/math/stats.

I mean no offense to the guy who got hired... I am happy for him, he is a friend after all but it just bugs me that the idiot HR employees who """work""" there hired him and not me when I applied a couple of months ago just goes to show that HR people are useless and braindead

Haha incredble Had this interview at bank... they didnt want me without explanation I know a guy .. very extroverted but incredibly incompetent with bad grades and absolutely no real practical knowledge in finance/stats/maths... only degree+lil work experience &they hired him😂

RIP John Hull

Crazy story: A bank analyst published a report showing Kalshi users lose money even faster than sports gamblers, so Kalshi first tried to pressure the data provider to change their data and then accused them and the analyst of conspiring to extort them when that doesn’t work.

With a few exceptions, of course ... such as a shortage of staff in a sector or company

Lesson learned: it doesn't matter how much self-studying you do ... if the interviewers don't like you or the way you talk (for whatever reason) you won't get the job

??? is this fake ???

@annanay also: love your content! Thanks for posting this!

@annanay Don't forget that random forests also use feature subsampling in addition to bagging

A single decision tree can learn something like “IF spread is wide and vol is high THEN imbalance matters”… but one tree is fragile. Small data changes can flip early splits and you get a totally different model. It’s unstable - sound familiar? In linear models, we fixed this with a technique called ‘bagging’. Train a lot of models, evaluate all of them, and average the result. When we do the same with trees, we make a ‘random forest’: data ├─ bootstrap #1 -> tree T_1 ┐ ├─ bootstrap #2 -> tree T_2 ├>pred_future_price (bps) ├─ bootstrap #3 -> tree T_3 ┘ └─ ... -> tree T_B Stepping back though, this seems kind of wasteful? Each tree is making some step towards a meaningful pred_future_price. When we fit the next tree, we lose all that work. How about we just fit each new tree to what the current forest is still getting wrong? Start with: pred_0 = mean(y) Then make the nth decision tree on the data {y - pred_{n-1}}. --> pred_n = pred_{n-1} + eta * T_n This technique of creating many sequential trees, each correcting the last, is known as boosting - some of the most popular implementations are LightGBM and XGBoost. More on them to come!

@systematicls I'm not really quant (mostly just interested in the methods etc.) ... but I'd be interested to join if possible

Given that we probably have one of the largest communities of professional quants in discord right now. I'm considering renaming our discord to "OnlyQuants".

A newly created Polymarket account invested over $30,000 yesterday in Maduro's exit. The US then took Maduro into custody overnight, and the trader profited $400,000 in less than 24 hours. Insider trading is not only allowed on prediction markets; it's encouraged.

How it feels running a small local LLM on your laptop

@__paleologo I like understanding deep learning, since a few people have already posted links I am also going to provide one udlbook.github.io/udlbook/

Observation: there are a few good foundational books about ML/AI: Bishop & Bishop: Deep Learning Fleuret: Learning Theory from First Principles Hardt & Recht: Patterns, Predictions and Actions and yet, none of them covers transformers in detail or at all. Question: can you recommend good grad level books on transformers and later DL? Add in replies.

Warren Buffett: “The secret to life is weak competition…” “‘How do you beat Bobby Fischer?’ You play him at any game except chess.”