As artificial intelligence continues to advance at a rapid pace, University of Nebraska-Lincoln’s computing courses are similarly evolving to prepare students for an AI-driven future. Courses that previously focused on foundational topics have been updated and expanded to combine classic computing principles with cutting-edge technology.
Data Analysis CSCE 320 has been offered in the School of Computing for several years, but has been redesigned to keep up with new technological advances and shift focus to modern methods. To align with the new direction, it will soon be renamed “Data Analysis with Machine Learning.” Students in the course, taught by Professor Ashok Samar of the School of Computing, learn how to analyze data using the latest algorithms, machine learning tools, and hands-on approaches.
“Two years ago, we completely changed our content and moved from being traditionally database-centric to machine learning-centric,” says Samal. “There are courses that teach you the basic fundamentals of machine learning, but this is more of a hands-on experience. We focus on how to use machine learning techniques, what machine learning techniques are, and best practices.”
Samar said that unlike many other computing courses, CSCE 320 does not require advanced coding experience because the curriculum focuses on leveraging tools rather than creating them.
“We have a lot of programming challenges, but they are more about how to use existing APIs than starting from scratch,” Samar says. “They still need to code. Rather than trying to code machine learning tasks, they learn how to use APIs that are already available in all kinds of packages to solve the tasks.”
Although this course is required for data science majors, its focus on application rather than development makes it open to students of all majors and especially valuable to non-computing majors in adjacent fields such as mathematics and engineering. Results from the first assignment, which investigated the majors of students enrolled in the course, showed that about one-third of the students in last semester’s class were non-computer majors.
“Machine learning is AI, and it is pervasive in our daily lives, so it is important to all fields, not just data science and computer science majors,” Samar said. “I think there will be interest in different fields because they all have large datasets and are looking for techniques to understand how to use them and perform machine learning tasks.”
CSCE 320 students can expect to explore a variety of topics through assignments and learn how different techniques can be applied to a wide range of fields. Examples of additional datasets include medical images of tumors, penguin types, sleep cycle lengths, IMDB movie details, Bob Ross paintings, and more.
The variety of subject matter reflects the School of Computing’s Data Science major, which is uniquely structured to be an interdisciplinary major. Students are encouraged to select additional areas of focus in a second major or complementary subjects that strengthen their data science knowledge and apply it to another area of personal interest.
“As long as you have large datasets and are willing to leverage those datasets to understand different aspects of problems in your field, it definitely opens doors for people from different fields,” Samar said. “Whether it’s sociology, psychology, art, or engineering, machine learning tools can help.”
Coursework with interactive ZyBooks lessons, lectures, in-class quizzes, and labs includes classification, clustering, cleaning, visualization, and evaluation. Students will learn how to identify appropriate models, select appropriate software packages, evaluate results, and adjust models as necessary.
“They train the model, test it, and if it doesn’t work, they go back and adjust the parameters or try a different technique,” Samal says. “Evaluation and refinement is a very important part of the course.”
Samal said an equally important element is understanding how to correctly apply concepts when using tools to produce and ensure accurate results.
“With all these tools, it has become equally important to understand how to frame questions to get the information you need,” says Samal. “Asking the right questions can take skill.”
This course curriculum will leave many students well-prepared for future data-centric careers, while also providing additional professional opportunities. It is an option to complete the IBM Professional Certification through Coursera. Students who choose to complete the certificate will receive both the resume certification and the final exam point deduction at no additional cost.
“I hope it’s useful and something I can post on LinkedIn, but it’s also complementary because there are some topics that I don’t cover in as much detail,” Samar said. “They can cover these concepts further and demonstrate that they have a professional certificate.”
Samar said that by mastering these concepts and learning how to analyze not only datasets but the machine learning problems themselves, students will be able to discover solutions and adapt to changes, no matter how technology changes in the future.
“You can solve a lot of machine learning problems using the tools that are available today, but it’s not the tools that are important,” Samal says. “It means they understand the core problems of machine learning and how it can be leveraged to solve real-world problems.”
