DS Analytical Execution\n地里的老题目了，大致如下\n日活用户的平均评论变多了， 因为机器人账户发了更多的评论。简单计算有多少机器人。然后讨论了怎么检测机器人。怎么去衡量是否成功，以及有什么影响。. Χ\n\nDS Technical Skills\n不确定地里有没有这SQL题。关于店铺的是否对客户可见。题设是，店家的店铺可以随时切换店铺是否对客户可见。\n\nDS Behavioral\n松弛的聊天。. ----\n\nDS Analytical Reasoning\n地里的老题目，在Post里加入餐厅推荐。\n. 1point 3 acres\n补充内容 (2024-12-11 03:10 +08:00):\n\n更新：12/6 拿到拒信\n\n\n更新面试题细节：\n有关daily active user (DAU)的经典题目：\n\n1.先猜一下DAU的#comments应该服从什么分布，我猜了normal。别问为什么，问就是normal比较简单。\n2.假设从所有的DAU里抽取两个1%的sample，请问这两个sample的#comments的mean的差值的95% confidence interval是什么？\n这题我没做出来，但答案可以问chatgpt，如下\n\nDAU一般的日均#comments在5-6左右，突然有一天DAU整体的#comments涨了20%。为什么？因为有bots。已知bots的#comments在500左右，DAU有2M左右，请问有多少bots？\n简单数学计算\n怎么去标记bots？\n怎么衡量是否成功？\n如果标记bots的算法FP过高会怎么样？\n用哪些feature来标记bots？\n\n\n\n关于上线商店的问题，table如下\nshop_visibility\n+------------+-----------+-----------------------------------------------+\n| column | data_type | description |\n+------------+-----------+-----------------------------------------------+\n| profile_id | BIGINT | Unique ID for a shop profile |\n| date | DATE | Day of visibility update in yyyy-mm-dd format |\n| ts | TIMESTAMP | Time of visibility update |\n| visibility | INT | Visibility: 0 (disabled) or 1 (visible) |\n+------------+-----------+-----------------------------------------------+\n\n\n\nWhich are the 10 shop profiles that changed their visibility status the most?\nWhat percentage of shops are currently visible?\nHow are currently visible shops distributed across the different categories?\nWe believe that over time, our new user base has shifted from profiles that had their shops visible most of the days to shops that are only visible for very few days. How could you try to verify this with the data given to you?\n\n\n\n\n\n\n补充内容 (2024-12-14 05:56 +08:00): ..\n\n谢谢小伙伴提醒。加上表格\n\nshop_details:\n+---------------+-----------+---------------------------------+\n| column | data_type | description |\n+---------------+-----------+---------------------------------+\n| profile_id | BIGINT | Unique ID for a profile |\n| shop_category | VARCHAR | Can take many values |.google и\n| visible_days | BIGINT | Number of days shop was visible |\n| creation_date | DATE | Date on which shop was created |\n+---------------+-----------+---------------------------------+. 1point3acres.com\n