Post

NUS: Year 3, Semester 1 (SEP @ PSU)

Navigating New Waters: Y3S1

As I set sail on my third year at NUS, I found myself on a new adventure—the Student Exchange Program (SEP) at Pennsylvania State University (PSU). This semester, I successfully mapped three computer science courses, one UE, and one Gen Ed course. The overall experience was enriching, with the academic rigor being noticeably less intense compared to NUS. I even managed to achieved a spot on the Dean’s List for the first time—a feat I had not accomplished back at NUS.

The most striking difference was the necessity to attend lectures—a practice I had not followed back at NUS. This shift in academic habits was quite a change, but it was part of the unique experience. Beyond academics, I had the opportunity to explore both the US and Canada (where I spent Thanksgiving break), broadening my horizons significantly.

While you can refer to here for my SEP experience or here for my tips and tricks for course mapping and SEP application, this post will primarily focus on the courses I took, their mappings back to NUS, and the differences I observed between the two institutions. Stay tuned as I share my journey of navigating these new waters.

CMPSC 448: Machine Learning and Algorithmic AI

Mapped Course

CS2109S: Introduction to AI and Machine Learning

Course Review

Overall, the course had a relatively low workload. However, my semester differed significantly from previous ones as both midterms and finals involved comprehensive ML projects. The midterm project focused on Part-of-Speech (POS) tagging and required group collaboration, while the final project offered the flexibility to work individually or in a group and to select a topic of choice. Nonetheless, the project mandated the use of two out of three models: CNN, RNN, and transformer-based systems.

The course content primarily covered the mathematics behind ML models and provided explanations for these models, but it was loosely connected to the projects. Therefore, it’s crucial to be well-informed, conduct personal research, and have reliable teammates. I opted to undertake the final project solo as I wanted to challenge myself, and managed to received full credit, making the effort worthwhile.

Lecture content was directly relevant to the four quizzes conducted throughout the semester, but did not test all the complicated math formulas, so it was pretty easy. Most people scored 9/10 for each quiz.

Performance

I performed exceptionally well, achieving approximately 96%. It appears that grading at US institutions is more lenient, and there was no grading curve applied to this course.

Grades

  • Expected: A
  • Actual: A

Comparison to NUS

Initially, CS2109S at NUS was quite challenging when first introduced by Ben Leong. Since then, the course has undergone several modifications, leading to a significant increase in difficulty. Despite these changes, the workload remains substantial. Nonetheless, CS2109S continues to be a popular choice among students looking to map their courses during an exchange program.

CMPSC 443: Introduction to Computer and Network Security

Mapped Course

CS3235: Computer Security

Course Review

The labs and homework assignments were quite time-consuming, with five labs and six homework assignments taking up the majority of my time. The difficulty was average, but it definitely took a toll. Lecture attendance was mandatory, tracked through Canvas quizzes given at the end of each lecture, which required a password. The lecture content was extensive, delving into the inner workings of various topics. I found the section on cryptocurrency particularly interesting and enjoyable, despite the breadth of content.

The finals were closed book, with straightforward and simple questions that essentially tested memorization. Having a strong foundation in networking, operating systems, and cybersecurity concepts, I managed to study just two days before the finals and still performed well.

Performance

I completed my finals in an hour and was ranked above the upper quartile with the score of 84.

Grades

  • Expected: A-
  • Actual: A

Comparison to NUS

The structure of this course can vary significantly from one semester to another. My friends have previously complained about the intense workload associated with the project component. However, for the current AY23/24, there were more quizzes and assignments in place of the project. If you enjoyed CS2107, you might consider taking this course. CS3235 heavily overlaps with CS2107, similar to CMPSC 443 at PSU. For those specializing in information security or have Computer Security as their Focus Area for CS majors, this course serves as a prerequisite for all other cybersecurity courses (except IFS4101), so there is no escaping it. However, it is advisable to check which professor is teaching in each semester and plan your schedule accordingly.

CMPSC 410: Programming Models for Big Data

Mapped Course

CS4225: Big Data Systems for Data Science

Course Review

CMPSC 410 was divided into two classes, each with different components and tests depending on the professor. The first half of the semester, including the midterms, was relatively straightforward. We primarily used Python to implement MapReduce, and the main challenge for most students was mastering Vim—an obstacle that NUS students, accustomed to the rigor of CS2030S, would likely find manageable.

However, the second half of the semester proved to be much more demanding, particularly the labs. We also had to learn Scala, which added to the complexity. The content increased dramatically, with more than half the semester dedicated to simple MapReduce concepts and less than six weeks allocated to Spark. The final exam consisted of only three practical questions to be solved in two hours. As practical tasks are not my strong suit, I managed to fully solve only two of the three questions.

Performance

I achieved full marks for the midterms (30%) and the labs. However, my performance in the finals (30%) was average, resulting in a final grade of B+. Considering the class’s difficulty, I estimate that fewer than five students received an A.

Grades

  • Expected: A-
  • Actual: B+

Comparison to NUS

I would definitely recommend taking this course at NUS. It is commonly regarded as the easiest level-4000 course, with all coursework in Python and Java. The finals are effectively the only differentiator since most students achieve full marks for the assignments. Only reason I took it on exchange was because I had no other CS courses to map, and I did not need to take any more UE.

CED 152: Community Development Concepts and Practice

Mapped Course

CLC1101: Engaging and Building Communities

Course Review

This course fulfilled my General Education (GEN) pillar requirement. For a list of non-GE-coded courses that fulfill GEN pillars, you can refer to here.

CED 152 was an engaging course that I thoroughly enjoyed. It provided a deep dive into the concepts and history of community engagement within the US context. The workload was slightly higher than anticipated due to the absence of two team members. The readings were extensive, and while I wasn’t very consistent, it’s crucial to note key points as in-class quizzes often test on these readings, particularly the textbook.

Performance

I achieved full marks for my assignments and received full credit for in-class quizzes. The project component included a preliminary report, a book report, and an in-class presentation, which made up a significant portion of the final grade.

Grades

  • Expected: A-
  • Actual: A

Comparison to NUS

I would definitely recommend mapping CLC1101 during an exchange if possible. Securing a spot in CLC1101 at NUS is becoming increasingly challenging as more students realize that taking one of the three CLC courses can fulfill the GEN pillar—a much more attractive option compared to the current GEN courses. I would advise mapping CLC-coded courses rather than GEN-coded ones due to the nature of our mapping system and the context of GEN courses. Mapping requires content overlap, and considering that GEN courses often include Singaporean history, finding an equivalent course overseas might be challenging, even if the underlying concepts are similar.

ASIA 177: Rise of Modern Southeast Asia

Cross-listed with HIST 177: Rise of Modern Southeast Asia

Mapped Course

HY2250: Introduction to Southeast Asian History

Course Review

I absolutely loved the course. The content was very accessible, partly because some aspects of Southeast Asian history are familiar to us as Southeast Asians. However, I also realized how much I didn’t know about the region’s history, perhaps because our secondary school social studies/history lessons didn’t focus much on neighbouring countries. The material was easy to digest, and the professor was incredibly engaging and welcoming, especially to me as an exchange student. Her teaching style fostered a warm and inclusive atmosphere in our small class of about 29 students.

Performance

I scored full marks for nearly everything. The surprise in-class quizzes were straightforward, requiring just a bit of effort in memorization. The professor even allowed you to change your answers for certain quizzes before submitting. There were six essays throughout the semester, but you only needed to complete at least three to get full credit, as the professor would take your best three essays. She even offered a chance to revise your essay if you didn’t score full marks initially, allowing you to improve based on her feedback. The final was an oral exam on a country of your choice, which required more memorization, but it was quite relaxed, and I believe she was generous in grading as long as you showed effort.

Grades

  • Expected: A
  • Actual: A

Comparison to NUS

I don’t have any comments on the comparison to NUS for HY2250, as I don’t know anyone who took it.

Academic Summary

Honors: Dean’s List

Grade Distribution

GradeNumber of Courses
A4
B+1

GPA Analysis

  • GPA in PSU: 3.87 / 4.0

Given that the courses taken during the exchange program are graded on a pass/fail basis in NUS, my cumulative GPA at NUS remains unchanged.

This post is licensed under CC BY 4.0 by the author.