申请季基本结束了,这个过程和结果和预期的完全没有任何相似的地方。

先说结果,Fall’22申请了11所学校15个项目,基本都是MSCS及相关项目,最后1 admission, 1 waitlist, 13 rejections

[Oct. 2 Update] Spring’23 重新申请了MSIN@CMU,被录取了!感谢CMU,感谢INI,感谢Jennifer。历史大半年的申请季终于超级圆满的结束了~

BG

陆本(众多top3中给分最烂的那个),3.2 109(S25) 326(3.5),奖学金若干,ACM金若干,pub共一一篇,两段大厂SDE/MLE实习共计一年

找工向,排名敏感,费用和loc不敏感

申请结果

申请前 预估的bar(轻喷)分

彩票

  • MS CIS@UPenn (Dream school,实习leader之前读的项目,本来不了解,但一直跟我说要吹UPenn,加上有Alumni的推荐信就申了,但非高三维就是送钱)
  • MSIN@CMU (没想到能waitlist,后来分析了原因,放在之后的版面)5.7更新:WL->Rej
  • MCS@UIUC (申请前地里帖子消息说基本不招陆本,就业数据好,bar高,想着买彩票,但今年出现扩招情况,bar跟MCS@Rice持平(?),但还是没中奖)
  • MSCS@Columbia (听说人多又水,当然只是申请季前的看法)

主申

  • MSSE-SV@CMU (deferred to Spring 2023) (n连拒之后的救命稻草。听我说,谢谢你,因为有你,温暖了四季)
  • MSCS@Duke (导师原来是Duke教过书,以为有优势,没想到这个项目这么小,有眼不识泰山…)
  • MSCS@UCSD (今年缩招)
  • MCS@Rice (bar越来越高了)

保底

  • MEng CS@Cornell Tech (一年项目,以为能保底)
  • {MSCS, MSCS-AI, MSCS-DS}@USC (所有人都跟我说能保底,但没人说过硬卡gpa啊…纯纯是做慈善)
  • MSCS@NYU {Courant, Tandon} (bar越来越高了,tandon劝转applied urban science)
  • MCS@UCI (bar越来越高了)

其他想申但是因为推荐信数量limited没申的

  • 彩票类(Stanford, UCB, UMich, UMass, Cornell)
  • 主申类(JHU, UChi, Northwest, Brown)

这个申请季犯的错

  • 眼高手低,严重低估GPA在申请中的作用。由于是头部985+软背景还不错,于是选校及其激进,而且也不舍得放低身段了解排名靠后的学校(甚至是NEU,BU这种),导致几乎全聚德的局面。当然,到后期等不到ad的时候已经悔不当初,差点加申了。
  • 没有对各个program仔细地研究。片面的只了解部分MSCS项目,而且只通过自己对学校的印象和网上只言片语的评价选,没有考虑ma不match以及每个项目的特殊性(招工数据、卡GPA、招生人数、陆本bar、名校友好…)。事实上,申请季快结束的时候我才了解到一些对我本科友好的一些学校:JHU和UChi。SE-SV是实习leader告诉我的,不然真的会面临失学。这个是很多独立申请的人都会犯的错,建议是可以加一些申请群看下别人的讨论。
  • 信息滞后。今年一些项目的bar上涨的速度超过了预期,甚至20、21年的dp都已不适用,传统的保底校已经不再能保底。但是感觉这个也没法预期,很多人也没料到这种情况。从我上面的项目分级就可以看出有多么搞笑。
  • 中介服务不到位。也不算犯了错,但是花了5W觉得没有享受到等值服务,无非就是害怕信息差,但最终中介也没有提供有价值的信息,都是自己了解到的。就是给你填填申请表(但大部分都是我亲自填的),文书基本也是我自己重新洗了再拿到Fiverr改,整套服务感觉价值不到5000。

申请CMU的经验

  • 三围似乎真的不如其他学校重要
  • 文书很重要。MSIN和MSSE-SV都有小作文,我MSIN的5个问题几乎花了一整天,我觉得没有被直接拒很大一部分原因是这部分写的比较好
  • match程度很重要。建议从项目主页上找信息看如何写满足项目的要求,比如MSIN我着重强调了system的经验和business insights
  • 可能比较看重工作/实习经验。这是以往的经验,招了很多2+年工作经验的,放在我的case也是符合的
  • 分享一下MSSE-SV的小作文

Demonstrate an understanding of what software engineering is, your familiarity with the specific goals of the program, and your interest in the program’s content.

Please limit text to no more than 300 words


Software engineering, for me, is the process of designing, developing, testing, experimenting, maintaining and iterating large-scale functional programs. My understanding of software engineering largely comes from my intern experiences, especially at Bilibili, where I supported its recommender system. It is not difficult for me to develop separate modules in a simple course project owing to my 7-year programming contest experience. However, I had never fathomed that developing under industrial guidelines was far more complex.

There are several significant elements contributing to the success of a software project. To understand the user or client requirements and organize the development process is essential to cultivate valuable outputs in the first place. Practice methodology also plays an important role in software engineering. For instance, adopting the SCRUM framework can exploit the team’s ability to respond to emerging requirements and deliver quickly. Industrialization is also an aspect that should not be neglected: Data, AI middle office can significantly leverage the developing efficiency. Even making a simple tool by yourself sometimes can help avoid lots of dirty work. What is more, requirement review, code style, documentation, etc., contribute to software engineering in different ways as well.

The sophisticated nature of software engineering has really fascinated me. It requires not only technical skills but also a deep understanding of ultimate goals, practice methodology, architecture designing, cooperation and so on. Only who mastering all these skills can be a good software engineer and make successful products. I really wish to go further in this field in the future. Therefore, I am particularly attracted to the Master of Science in Software Engineering program at College of Engineering, Carnegie Mellon University, which will definitely consolidate and build upon my ability to solve real-world problems and strengthen my practice skills, and will play a significant role in my long-term future as I seek to build influential products to help understand and improve our society, and achieve success in my career.

虽然过程比较煎熬,但这个结果最后看起来也非常不错了(录得多不如录得巧)。根据@Flash的电子云理论和倒三角选校(ref: https://www.1point3acres.com/bbs/thread-883934-1-1.html ),我能拿到还不错的结果应该也是all in彩票的一个极端长底短高的三角。