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
最近看了很多地里的帖子,自己也在捉摸着开始为未来跳槽做面试准备。
楼主目前location在波士顿这边。投了一家叫做DataRobot的小公司贡献给大家点面经。
这家公司主要做machine learning platform提供给DS或者DA进行数据分析。
它家除了在波士顿有个点,貌似在欧洲还有其他地方也有。
面经流程比较奇葩,有大概4,5轮电面。因为基本上每次面试官都是乌克兰那边打过来,所以我都挑的早上八点跟他们面。HR说如果5轮面完去onsite还有两轮,但貌似是behavior面。
前面几轮基本都是coding和machine learning的经历面,最后一轮是design面设计uber,一直是他提问我回答问题。因为楼主之前找工作也没怎么面过design,所以可能准备的不充分被design挂了。
anyway,我把面的一些问题都记下了,没有按顺序打出来。coding都很简单。面了感觉他家有的工程师水平一般,写对的代码以为写错了还是我提出来才看懂。。
1.
# Write a function fib(n) that will generate the nth number in the Fibonacci sequence.
2.
# Suppose we have a function that returns a list. Write a function that returns True if every element
# in a reference list is also in the returned list (order is not important).
# For example:
# If the expected value is: [ 1, 2, 3, 4]
# and the actual value is: [3, 2, 1, 4]
# The function should return True. However, it should return False if the actual value was:
# [1, 1, 2, 3, 5]
# The function should work for any two lists, not just those used in the example above.
3.
# Given a CSV file with three columns ‘user_id’, ‘item_id’, and ‘rating’:
#
# Example data:
# his is equivalent to [0, 56, 0, 0, 0, 91, 0, 0, 12]
# a2 = [[0, 23], [5, 43]] # this is equivalent to [23, 0, 0, 0, 0, 43,]
#
# print add_sparse_arrays(a1, a2)
# [[0, 23], [1, 56], [5, 134], [7,12]]
5.
# Given a CSV file containing user_id, item_id and rating, find the median rating
# for each user and then print out every item_id having a rating equal to the median
# for the corresponding user.
# Example data:
# user_id,item_id,rating
# 1,1244,2
# 1,62212,1
# 1,19394,4
# 2,912,3
# 3,1511,3
# 2,913,4
# 2,1939,1
# 1,5234,2
# 4,764,5
# etc.
6. design uber
一些general的问题,我就把scalability里那一套东西拿出来说了说。
|