注册一亩三分地论坛，查看更多干货！
您需要 登录 才可以下载或查看附件。没有帐号？注册账号 
x
都是十月份面的，拿出来回馈一下题目都是直接从我自己面完自己记录的doc文件里直接复制粘贴过来的，可能有公式的源码，大家凑合着看吧, 当时用英文记录的，就不帮大家翻译成中文了
Two Sigma, Quant Researcher:
OA: 地里的老题，用Pandas处理数据还有temperature linear regression之类的
Phone: 问了问简历里面的ml和optimization的东西，然后两道题
(1) daily log return with mean \mu and standard deviation \sigma. What’s the annualized Sharpe Ratio? If the daily log return has an autocorrelation of \rho, what‘s the Sharpe Ratio if we use the weekly cumulative return (what’s the numerator and what’s the denominator)? How the variance will change with \rho?
(2) Model the New York House Price with two variables: total area and number of bathrooms. What’s the sign (positive / negative) of these two variable in a linear regression model do you think? When you really run the linear regression, do you think it will be what you expect? Why? How to deal with multi-collinearity? What about regularization?
Onsite (3 rounds)
(1) Merge intervals: given intervals [[x_1,y_1],[x_2,y_2],...,[x_n,y_n]], merge all overlapping intervals into one large interval
(2) Humidity Query: given humids = [['city','time',humidity']] and queries = [['city','time']], output all the corresponding humidity of each query such that the query time is closet to but no earlier than the corresponding one in the humids.
(3) Ridge regression, how to tune hyperparameter, what does it imply if the out-of-sample performance never change however we tune the hyperparameter?
(4) Two random variable X_1,X_2 are jointly normally distributed with mean 0 and covariance matrix [[1,0.5],[0.5,1]]. Given one observation that x_1=0.5, what’s the estimation for the corresponding x_2? What’s the 95% confidence interval for your estimation?
(5) Assume we have n random variables which are jointly normally distributed with mean 0 and cons? What if we pick n balls? (bar insertion method / integer programming method)
(3) repeatedly roll a die until the sum goes over 63. what’s the distribution of the last num that doesn’t go over 63 (i.e. 58, 59, 60, 61, 62, 63)?
电面2：
(1) What’s Random Forest? Difference from Decision Tree? What’s Bagging? Why random features could reduce the variance?
(2) Assumption of Linear Regression? Will multicollinearity bias coefficient? How Will heteroskedasticity bias coefficient? What if residual follows t-distribution?
(3) Linear Interpolation: Given a sorted array A and target x, return linearly interpolated value of x. Follow up: time complexity O(n) to O(logn)
e.g. A = [[0, 1], [2, 4], [5,7], [8,6]], f(0) = 1, f(1) = 2.5, f(7) = 6.33
(4) Merge two sorted arrays inplace
A = [1, 2, 3, 4, 0, 0, 0], m = 3, B = [2, 5, 6], n = 3 f(A,B,m,n) = [1,2,2,3,4,5,6]
(5) Pandas manipulation
(1) mean of certain feature grouped by name
(2) weighted mean of certain feature grouped by name
Onsite:
现场做Data Analysis Project。 做到一半收到另一家公司offer，就把这onsite给鸽了，果然还是败人品，劝大家以后千万别做这种事，我还是太天真了。
以上，都是onsite面挂了，或者面过了但是最后种种原因没拿到offer的，发出来祭奠秋招
|
匿名用户-QT1UT
2020-12-18 03:09:52
狗仔卡
收藏
淘帖
好苗
杂草
提升卡
置顶卡
关贴卡
开贴卡
变色卡
千斤顶
楼主
