Page 83 - Special Topic Session (STS) - Volume 3

STS515 Jeremiah D. D. et al.
will be helpful for us to derive a thin-core undergraduate DS curriculum, which
may be flexibly augmented by electives in senior years, including some
strongly domain-related courses such as bioinformatics and computational
finance.

5. Discussion and Conclusion
In this paper, we have surveyed a number of representative data science
undergraduate curricula and revealed some intersting diversity between these
programmes. Perhaps the inherent interdisciplinary nature of data science
itself justifies and demands its diversity, and hence it will be futile to keep a
consistent curriculum design. Rather, institutes may find it more rewarding to
install a thin-core but multi-facet programme that accommodates students’
and employers’ diverse interests.
This flexible design implies running classes with strong diversity and poses
new challenges to mathematical teaching. The strong correlation between
students’ performances in mathematics and engineering subjects has been
confirmed (Bishof et al., 2015). On the other hand, Ooi (2007) criticised the
usual, administratvely efficient mode of mathemtics teaching delivered as
separate subjects, resulting in low relevance perception among Engineering
students. Are these findings relevant to DS as well? This is a topic to be
investigated in our future work.
Table 3. Top-10 data mining algorithms and their corresponding
prerequisites as required.

Algorithm MATH100 MATH200 STAT100 STAT200
C4.5 x
k-means x
SVM x

Apriori x
EM x x
PageRank x
AdaBoost x x
kNN x
NB x

CART x
# Ticks 4 2 4 2

72 | I S I W S C 2 0 1 9