Data Science之你問我地答

岩呀
識Linear regression又話自己識
識call lib做cnn又話自己識
根本冇baseline

過hr真係最難果關

留名，有無巴打睇過The elements of statistical learning？睇左20幾頁已經覺得好撚難

some of the IMPORTANT derivations in this book are put in the exercises
for example, the derivation of smoothing spline

what you need is ... a solution manual written by others...
https://waxworksmath.com/Authors/G_M/Hastie/WriteUp/Weatherwax_Epstein_Hastie_Solution_Manual.pdf

數底差搞到睇得好辛苦，d equation 又成日跳step

Thx巴打，我都搵到呢個source，依家都係一路睇一路google

何苦呢

係咁㗎啦，佢係得登㗎，逼你要拎支筆出黎諗吓嘢，唔係只係齋睇。Maths textbook半頁睇一日唔係乜嘢特別事
如果真係想明晒同記憶深入啲，係應該自己慢慢complete 返嗰proof, 唔係好難領悟倒啲人點諗倒個proof 㗎。

我讀master 嗰時會將啲non trivial but important 嘅proof次再寫低做note, 加好多annotation 去提一個step點跳，而家都成日會refer 返啲note

例子

無computer science底但想學algorithm/machine learning
點入門?

Algorithm既話
如果自問math OK可以睇下Kleinberg同Tardos本Algorithm Design
math唔OK既可能睇下discrete math先

https://t.me/DybalaJuve10
巴打請加我，想一齊學習交流🙇🏼‍♂️

look up "double decent" in neural networks if you haven't

又或者睇An Introduction to Statistical Learning打下底
呢本淺好多, 不過之後睇ESL, 數嗰方面應該都係要靠自己

Stanford Statistics 315a Lecture note
http://statweb.stanford.edu/~tibs/stat315a/

maybe useful...

有冇巴打做緊/過/ 識人做馬會data scientist? 分享下

唔同department 應該會請唔同要求既DS

主要想睇下工作氣氛如何，tools 先唔先進

大公司應該每條team唔同啦

Thx巴打,等我研究下先

考試真係記憶力大賽，明咗易記啲，不過overall都係難記，尤其考試一次過要背10條proof嗰陣時。

呢啲睇prof la, 我嗰時讀嗰個有得帶一張A4雙面cheat sheet （其實用唔著多，prepare chests sheet 只係温書嘅一部分）同埋要你即場proof 嗰啲都係啲一一定要識嘅basic 嘢，唔似assignment 嗰啲咁又諗餐死...

想問python/r，識到咩程度先算係「識」
而家係咪基本上無公司會用SAS？

另外有無巴打黎緊9月會讀DS master，可以相認下

有唔少傳統既大公司仲用sas，但會越黎越少
起碼識晒data import, featuring engineering, model building validating掛