[UMN MSBA] Some Reflection on Courses in Fall Semester
Right now, I am waiting relaxingly in MSP, ready to board the plane heading to LA. I try to conjure up and summarize this semester as much as possible. It is my second semester studying at UMN MSBA. After the training at summer, the fall semester seems to be much more flexible and leisure owing to the length of term.
As expected, the courses become much harder, the team work becomes more challenging and the job hunt becomes more urgent. I was busy balancing different task at U, including the schoolwork, my internship and my career exploration.
We have 5 courses this semester. Each of them is the whole new field for me. And instructors led us to walk into the forest of analyses and showcased how huge the analytics world is.
I will try to briefly introduce what each course is about and just like what I did on summer course, give each course some scoring, aiming to grant those who are interested in UMN MSBA some clear descriptions and idea what I have gone through in this four months.
a. Data Management, Databases, and Data Warehousing
General Level: 6/10
Growth Level: 8/10
Breakdown Level: 5/10
The course is taught by the instructor who taught us Python in summer. Unlike Python course, instructor is more laid-back than in summer. In this course, we learn relational database language, alias MySQL, which is widely used in job market.
In the first two thirds of the semester, we learn basic and intermediate skills of SQL language, from select, join, subquery to window function and recursive. In the rest of the semester, we learn how to design conceptual database. The reason we learn how to design database is because according to our advisor board, it is another required and hot skills needed in daily task of a business analyst. Business analysts don’t have to build the database infrastructure, but they have to know how to communicate with engineers to co-build the desired structure for better data quality.
The course scoring consisted 8 assignments, several self-assessment, 3 exams, 1 SQL challenge and 1 final trend market presentation. I think after the class we still can’t be the SQL expert. Instructor put emphasis on different components of lectures quite unevenly. He spent too much time dwelling on the simple query, and thus leave us limited time to explore more advanced topics. Also owing to the ill allocation of schedule, we experienced several adjustment in class requirement, which sometimes is hard to follow.
To sum up, the instructor style is a bit different when we took his python class. We got to start our exploration in SQL, and I start to understand SQL is an important skill that needs consistent practice. However, it is really hard to find related practice online to polish SQL skills. And it will definitely be one of my topics in the following month to self-learn and self-progress.
b. Big Data Analytics
General Level: 6/10
Growth Level: 6.5/10
Breakdown Level: 9/10
This course granted me the most challenge in this semester. Big Data Analytics, taught by our program academic director, covered most of the big data tools in the business analytics field, including Linux, Hadoop, Pig, Hive, Spark and AWS (cloud computing).
The course is combined with lectures, in-class demos/labs, 8 hand-on assignments, two exams and the final trend market presentation. The course is like a twin course of SQL course. First, some of the tools we practiced resembled SQL language (like Hive and Pig), Second, the trend marketing presentations are combined into one presentation, which is one project, scored by instructors in two courses. We have to introduce one of the big data/database related tools and combined it with business use case. The presentation was set up in Carlson Atrium, guests came and go, we then conducted elevator pitch to introduce our technology.
The course is densely-packed, which is quite opposite to SQL course. There are too many topics to cover in the course. Sometimes, we have to absorb the contents ourselves, and then seek methods to review and finish the challenging hand-on homework. I think the main purpose for the course is to expose us to the world of big data. We are able to take a glimpse of parts of structure of the big data ecosystem, but what we are not able to do is to really link to real business world and put these techniques into practice.
This course is definitely the course I had tried my best to command in my MSBA life. It is not hard, to be honest, the assignments were similar to labs, and most of exam contents were the nuances elaborated in decks. However, it took some attention and focus to memorize those details for different tools.
Traced back to the origin of the challenge, student really needed to manage time well during this course. I tried to love this course and practice as much as possible, but it just didn’t turn quite right. I believe that for those excels at this course might proceed to the career path of data engineer. And those who aim at big names, having good commands of this course will be a good bonus, since some big names will indeed have infrastructure of big data tools such as Spark, Hive or Hadoop.
c. Exploratory Data Analytics and Visualization
General Level: 6.5/10
Growth Level: 9/10
Breakdown Level: 9/10
Exploratory Data Analytics and Visualization, alias EDAV, mainly focuses on one track of data science: unsupervised learning, where there is not specific features to predict upon. We have to use several algorism to explore the pattern of data.
The instructor gave us most of the lecture whose topics covered as many as topics of EDA, from association rules, clustering, dimension reduction to anomaly detection. The instructor is quite good at elaborating concepts and hard mathematical algorism, but there is still a big gap between understanding the concepts and application in the real business world.
The instructor aimed to make us polish our technical skills in the four project and final live case. The four homework projects needed to be finished in group. It was quite a disaster for these couple projects. My team usually got only half of the full mark. And the point which made it worse was that we never got what instructor and TA really wanted. For these projects, we needed to conduct the analyses with R, but I am not very familiar with this tool as I was with Python, plus I also had to spent a lot of time communicate and lead within group, which granted me lots of practice in leadership, and the high and ambiguous standard so hard to reach from the instructor. All of these led to some challenges.
I have to say, I learned a clear framework of unsupervised learning. I know clearly what kind of algorisms and techniques I need to polish and get more understanding of. I did polish some of them during the projects but not so much. And I believe until the last minute, I will never know they are useful or not.
d. Predictive Analytics
General Level: 7.5/10
Growth Level: 8.5/10
Breakdown Level: 7/10
Predictive analytics is the twin course of EDAV. It mainly focuses on supervised learning. It has a common name in the BA field: machine learning. Its structure is the same as EDAV: lecture as main stream, 4 group projects, and one final live case in predictive field.
The instructor, in my opinion, balanced quite well between lecture and practice. Some of the machine learning codes would be accompanied with his video recording, which is very helpful. But still, I sometimes felt quite lost in his lectures. The contents were sometimes overwhelming to me, owing to my lack of math and statistic background. I basically didn’t get understanding of neural network and the lectures afterward.
To briefly sum up, I think the instructor has tried his best to deliver as much knowledge as possible in the semester. But the world of predictive analytics is just way to big, and I was in the quite poor start point, which made me hard to catch up.
e. Live Case
Live case is a signature of UMN MSBA. For fall term, we have one company client and two live case threads: EDAV and Predictive. Our client is one of the biggest hospitality and entertainment company in Midwest America. For each thread, the problem statements are different. We started our analytics at the end of October and finished in the end of semester.
I think what I have learned the most from live case is how to see clear the reality and how to finish a project from beginning to end all way alone. And I really treasured the priceless experience. And only from the live case did I start to tell the difference in my classmates’ capability of analyses. Someone excelled and walked so ahead, and some just lagged behind so far.
Conclusion
This term is so fast. And the number I broke down is nearly the same number as I did in previous term. But I have to say I am glad I survive. I end this article during my road trip, and I believe my road trip in MSBA will continue. Stay tuned.
