Third-year first semester Introduction to AI:
The teacher will post each PPT; a sprint review a week before the exam is enough.
PS: Interested students can read Artificial Intelligence: A Modern Approach (4th edition) (Stuart Russell). (Teacher highly recommended edition) Professor Zuo said his agent chapter PPT was borrowed from this book.
Is that the 200‑person class?
Yes 
but later, after dropping courses, only about 100 left; the core foundation is mandatory, can’t drop it 
Junior Year First Semester Operations Research:
No extra study is needed; just follow Professor Liu Jia’s PPT. He will turn last year’s exam questions into regular assignments. If you do the regular assignments yourself, the exam won’t be any worse.
Junior year first semester Functional Analysis:
Sun Jiong “Functional Analysis” (Higher Education Press)
The whole book is rich in examples, easy to understand, and comes with many exercises. Although this book’s style is basically orthogonal to Professor Xu Xiaoxi’s exam style, it is indeed very suitable for the first pass of learning functional analysis (I only encountered this book in mid‑December, and I really wish I had it during the summer break).
Xu Quanhua “Lecture Notes on Functional Analysis”
This was recommended by a senior; he used it to review for the graduate entrance exam and it is indeed well written. When I self‑studied functional analysis in the sophomore summer, I used this book, but only got to the chapter on Hilbert spaces. Personally I felt it was a bit tough, so I didn’t continue (sigh, maybe I lack talent; I still think that for the first pass, a book with abundant, easy‑to‑understand examples is better).
A senior majoring in pure math said that if you only need the course‑level functional analysis, just read chapters 1, 2, 3, 5, 6.PS: Personal opinion: Professor Xu Xiaoxi uses his own PPT for teaching; before the midterm the material is relatively simple and tolerable; after the midterm, when the Hahn‑Banach theorem and reflexive spaces appear, the teacher often lectures while I’m trying to catch up… Juniors, remember to preview each class if possible.
Midterm scores (rumor)
2021 class: highest 43, average 20
2022 class: highest 35, average 15
Maybe the 2021 midterm was too easy (?) so the final was all new questions. According to Professor XXX, the exam was a disaster. This led to the situation that when grading, if a student wrote the correct theorem name, regardless of details, the grader would give a high score (15 points for anything above 10). The 2021 class was his first cohort. After learning from the previous year, before the 2022 final, Professor XXX said in class that the final would have 6 questions, of which 4‑5 would come from PPT/midterm/assigned homework/exercise‑class supplement (maybe slightly modified), and 1‑2 would be completely new. That turned out to be true. Perhaps the 攻略 is “midterm a bit weak, final lenient” 
Later: grades came out, about 10 points lower than expected 
Algorithms (Junior Year, First Semester):
I really didn’t grasp algorithms; I’m quite stuck. I have past exam questions; feel free to PM me if you need them.
The final scores haven’t been released yet; this is a draft the teaching assistant posted in the group. The exam felt a level harder than last year…
Junior year first semester Mathematical Statistics:
(This year it has been changed to teachers Yu Dalei and Li Xingxiang)
If the instructors are the two above, please pull yourself together, carefully review the teacher’s PPT, and stop thinking that “proof problems in mathematical statistics are fewer than calculation problems” The midterms and finals are truly difficult…
Junior year first semester Natural Language Processing:
Professor Jiang is a great person I feel he’s the best teacher this semester, the exam papers are also the simplest, just choose and you’re done (if you need past exam questions, DM me)
PS: As of today I haven’t written my NLP project yet, sigh, I’ll start tomorrow.
Junior year first semester Partial Differential Equations:
Junior year second‑semester Machine Learning:
In fact, Professor Meng Deyu follows the PPT, and the exam scope is also his PPT. Passing should be fairly easy, but getting a high score may be difficult. Grade composition = 60 labs + 40 final exam (at least 3 labs, no limit on number; submit an electronic report, topic self‑chosen, as long as it is related to machine learning)
[details=“This year’s recall version”]
Junior Year Second Semester Modern Mathematics Selected Lectures:
Professor Wang Bin teaches according to the PPT that corresponds to this book. The assessment method is to submit a course paper; anything related to the course content can be written about. There is no final exam, attendance has not been checked, grades have not been released yet, so it’s unclear whether the score will be “beautiful” (I feel Professor Wang is quite laid‑back, so it’s likely to be good).
Later: the mobile academic‑aff
Numerical Algebra (Junior Year, Second Semester):
Professor Wang Fei taught according to his own PPT, which differs from the textbook. The exam scope also follows the PPT. Grading composition = 60% assignments (4 × 10 assignments + 1 × 20 final paper) + 40% final exam.
Basically all the key points were covered, and the question types were quite flexible. I didn’t do very well, and grades haven’t been released yet. However, according to Professor Wang, this course will be cancelled for the next cohort, so I don’t really need to share any key points. I post this as a tribute to this 2‑credit course.
Update: It was indeed unsatisfactory; the score is at the lower bound of my estimated range 
Junior Year Second Semester Intelligent Perception and Mobile Computing:
The whole course is taught by two teachers, Huiwei and Wang Ge. Professor Hui mainly covers sensors and computer networking, while Professor Wang Ge focuses on the mobile computing part. The exam scope follows the teachers’ PPTs, and the grade composition = 30% regular (assignments + thought questions) + 70% final exam.
Exam paper composition: 30 multiple‑choice questions (15 × 2) + 40 calculation/short‑answer questions (8 × 5) + 30 long questions (15 × 2). There are no overly obscure or deep questions. The exam questions are not difficult; many students submitted their papers an hour early. Thanks to the teachers for being len
Third-year second semester Multivariate Data Analysis and Statistical Software:
Exam scope follows the teacher’s PPT (will include more content than the textbook). Grade composition = 50% regular (20 homework + 30 major assignment) + 50% final exam. I personally feel the knowledge points are quite fragmented, making comprehensive review difficult (maybe because my statistics‑related subjects clash) 
Multivariate Data Analysis:
Objective: fill‑in‑the‑blank questions (8 questions, 4 points each)
Long questions: calculations, proofs, and extracting information from R output.
Chapter 1: Descriptive analysis of data (at least 2 fill‑in‑the‑blank questions)
Chapter 2: Multivariate normal distribution (proof questions)
Chapter 3: Regression analysis (proof questions)
Chapter 4: Analysis of variance
Chapter 5: Principal component analysis, canonical correlation analysis
Calculator allowed.
Example 1: Build a logistic regression model (write its form); given a new observation, predict the response variable.
Example 2: Extract information from R output and interpret the results.
Differential Geometry (Junior Year, Second Semester):
The exam covers the entire book. Grade composition = 20% regular assessments + 80% final exam.
Exam format: 20 true/false questions (2 × 10) + 20 fill‑in‑the‑blank questions (2 × 10) + 45 computational problems + 15 proofs.
In class, the instructor said that the requirements for the foundational and major‑category exams would be a bit lower than for the math exam. Moreover, we were lucky this year
All exams during my undergraduate period end here, I declare this post closed! 