前言
新的學期開始，在美國也經過了一個quarter，放了一個長假，分享一下在芝大上了三門課的心得。
Program 介紹
我讀的program 是MS in Statistics，Program 有五門必修課，主要都是和統計知識有關，理論和應用都有涵蓋到，希望大家以統計為基礎，再依自己興趣走不同分支。另外還需要四門選修課，再加上一篇 Master Thesis。未來同學有一大部分會繼續讀 phD，也有部分同學會以在美國就業為目標。
到目前為止的感想就是上課步調非常快，作業loading 也非常重，再加上芝大學制是quarter 制，開始上課沒幾週就準備期中考，很常一交完作業就準備做下週作業，基本不會有空檔，雖然忙，但是訓練也很扎實。
修課心得
我 22 Fall 選下面三門課：
STAT 30030 — Statistical Theory and Methods Ia
STAT 34300 — Applied Linear Statistical Methods
STAT 33500 — Time-Series Analysis and Forecasting
前兩門是 Program requirement，後面時間序列是我自己選的選修。
STAT 30030 — Statistical Theory and Methods Ia
Instructor: Daniel Sanz-Alonso
課程主要是 Statistics 的介紹，從機率講到 Point Estimation, Hypothesis Test。課程 Note 整理的非常清楚，而且整門課架構清楚，主要內容就是大學部程度的數理統計，而且這門課本身就是 Undergraduate level，所以學習上沒有太多問題，作業都是計算題，考試也不會太難，唯一缺點就是因為老師個人因素，英文講不清楚，聽起來比較辛苦。
STAT 34300 — Applied Linear Statistical Methods
Intructor: Rina Foygel Barber
這門課相對於第一門，難度提升不少，內容主要就是從 Simple Linear Regression 講到 Multiple Linear Regression, Statistical Inferece，再加上各種延伸，包括：bootstrap, model selection, ridge regression, missing data 等等主題。老師的教學十分仔細、有脈絡，投影片清楚，但是語速偏快，對我而言，吸收比較慢。我很喜歡老師的教學風格，她希望我們可以真正理解模型的推論，而不是一直背東西，所以考試可以帶 cheat sheets。我能感覺的出老師想傳達的東西，但需要很多時間吸收理解，也算是讓沒正式學完統計推論的人，重新補齊知識。
除了理論講解，這門課也有很大部分的程式應用，語言是用 R，主要都是解各式 Regression 問題。期末有一個 final project，老師會給一組 data，要我們從頭活用整個 quarter 學到的東西做 data analysis，我覺得很棒，可以盡量活用所學。
STAT 33500 — Time-Series Analysis and Forecasting
Instructor: Jeffrey R. Russell
最後這門就是我上學期的選修，主要我是想補齊一些 time series 相關的內容，這門課比較特別的是他是 Booth 商學院底下開的課，這樣算體驗到全美頂尖商學院嗎？課程是用 Lecture 方式進行，內容包括 ARMA models, Vector Autoregressions (VARs), GARCH Models, DCC and Factor models, Cointegration 等等，基本涵蓋了 time series 大部分主題。老師上課口條清楚，講義也寫得不錯，複習起來整體性很好，作業每周都有，內容大概六成程式題、四成計算題，和同學一起討論不會太難，期中考和期末考都是從講義作業延伸，難度適中。
後記
讀完一個 quarter 後，深刻感受時間的緊湊，每個老師都把上課時間盡可能不要浪費，同學們讀書氛圍也很積極，只是我自己吸收的時間要再快一點，不然趕不上 loading 壓境，今天(1/3/23)新學期開學，希望一切順利，盡快找到自己真正想做的，make this journey worth。
Preface
The new semester has begun, and after completing a quarter in the U.S. and enjoying a long break, I'd like to share my experiences from taking three courses at UChicago.
Program Introduction
I am enrolled in the MS in Statistics program, which requires five core courses primarily focused on statistical knowledge, covering both theory and applications. The program aims to establish a solid foundation in statistics, allowing students to branch out into different areas of interest. Additionally, students must take four elective courses and complete a Master's thesis. A significant portion of my classmates plan to pursue a PhD, while others aim to secure jobs in the U.S.
So far, my impression is that the coursework moves at a very fast pace, and the workload is quite heavy. Given UChicago's quarter system, midterms arrive just a few weeks into the term, making it common to finish one assignment and immediately start on the next. There is virtually no downtime. Despite the intensity, the training is rigorous and rewarding.
Course Experience
During Fall 2022, I took the following three courses:
STAT 30030 — Statistical Theory and Methods Ia
STAT 34300 — Applied Linear Statistical Methods
STAT 33500 — Time-Series Analysis and Forecasting
The first two were program requirements, while the Time Series course was an elective I chose.
STAT 30030 — Statistical Theory and Methods Ia
Instructor: Daniel Sanz-Alonso
This course provides an introduction to statistics, covering topics from probability to point estimation and hypothesis testing. The lecture notes are well-organized, and the course structure is clear. The content mainly aligns with undergraduate-level mathematical statistics, which makes sense since the course itself is an undergraduate-level class. I didn't encounter significant difficulties in learning the material. The assignments consist of calculation problems, and the exams are not too challenging. The only drawback is that the professor's spoken English is unclear, making it a bit difficult to follow the lectures.
STAT 34300 — Applied Linear Statistical Methods
Instructor: Rina Foygel Barber
Compared to the first course, this one is significantly more difficult. It covers topics from simple linear regression to multiple linear regression and statistical inference, along with various extensions such as bootstrap, model selection, ridge regression, and handling missing data. The professor is meticulous in her teaching, with well-structured lectures and clear slides, though her speaking pace is quite fast, making it harder for me to absorb the material quickly.
I really appreciate her teaching style—she emphasizes truly understanding the reasoning behind models rather than rote memorization. Because of this, we were allowed to bring cheat sheets to exams. Although I could grasp the key concepts the professor wanted to convey, fully digesting them required a significant amount of time. This course was especially valuable for filling in any gaps in my understanding of statistical inference.
Beyond theoretical instruction, the course also involves substantial programming applications using R, primarily to solve various regression problems. The final project requires analyzing a dataset provided by the professor, applying everything learned throughout the quarter to conduct data analysis. I found this to be an excellent way to apply the concepts in practice.
STAT 33500 — Time-Series Analysis and Forecasting
Instructor: Jeffrey R. Russell
This was my elective course for the quarter, chosen to strengthen my knowledge of time series analysis. Notably, this course is offered under the Booth School of Business, so I guess I got a taste of studying at one of the top business schools in the U.S.! The course is conducted in a lecture format and covers ARMA models, Vector Autoregressions (VARs), GARCH models, DCC and factor models, and cointegration, providing a comprehensive overview of time series topics.
The professor is articulate, and his lecture notes are well-structured, making it easy to review and connect concepts. The weekly assignments consist of roughly 60% programming tasks and 40% calculations. Working on them with classmates made them manageable. The midterm and final exams are based on extensions of lecture content and assignments, with a moderate level of difficulty.
Final Thoughts
After completing one quarter, I've deeply felt the intensity of the academic schedule. Every professor maximizes class time without wasting a minute, and my classmates are highly motivated in their studies. Personally, I need to speed up my learning process to keep up with the heavy workload. Today (1/3/23) marks the beginning of the new semester—I hope everything goes smoothly, and that I can quickly find what I truly want to pursue, making this journey worthwhile.
English translate is generated by AI for reference.