The power of spectral techniques

It is quite amazing to think that we can solve \(Lx=b\) even faster than sorting the edges...

Puzzle

A puzzle given by an office-mate:

There is a polynomial \(p(x) = a_0 + a_1x + a_2x^2 + \dots+a_nx^n\) where \(a_0, a_1, \dots, a_n \in \mathbb{Z}^+\). Knowing only \(n\), you may query the value of \(p(x)\) for any \(x\). The goal is to recover the coefficients with minimal queries.

Cosmos: A Spacetime Odyssey

傳說中在 imdb 得到很高排名, 最特別的它是紀錄片! (紀錄片在 imdb 也可以 9+ 分一定不簡單)

source: wikipedia

很大程度上歸功於取代了一般紀錄片的 "某專家坐在 office 俾 speech 的場面" 而是用 CGI/片作為背景

P.S. host 坐的那 spaceship 非常似 Tardis, 而且經常出現 "travel through time and space", 很難不聯想起 Doctor Who 啊

吐槽

雖然說過「歐洲的城堡都一樣」，不過美國基本上沒有真正的城堡這點還是非常遺憾

還有很莫明奇妙的是，在 Espresso 旁的雪櫃有 85% 是 reduced fat milk，但同時又要落 syrup (罪大惡極)，耐人尋味

不過，至少還沒看到 canned coffee 這種不能稱為咖啡的物體.. 其實喝過「真正」的咖啡以後就完全鄙視像雀Ｘ的罐裝咖啡

但多得有 Amazon ，我早已在 Expresso 的路上奔馳

Quora haqathon 2014

Quora haqathon today, from 11am to 7pm - Pacific standard time! Features 9 problems mixed with tradition algorithm tasks, machine learning and system programming tasks.  Link to site.

Ontology
Linearize the tree - each query reduces to "in question q[x...y], how many of them start with prefix p?". Offline query + Partial Sum + Trie. Linear time.

Wombats
Maximum closure.

Labeler
Use training set to calculate \(\text{Pr}[q_i \in t_k | w_j \in q_i ] \) for all question \(q_i\), topic \(t_k\) and word \(w_j\). Improve using bi-gram.

Duplicate
Use \( \text{Pr}[w \in \text{question_text}_i \text{ and } w \notin \text{question_text}_j ]\) as classifying criteria - 60% accuracy. Consider also \( \frac{\text{view_count}_i }{ \text{view_count}_j } \) improved it to 70%.