Inspiration
Thai students are struggle not only knowing what they like to learn but choosing their faculty. Dropouts and unsatisfactory college journeys are common outcomes of students picking the incorrect faculty. Nonetheless, we strive to build a platform that allows students to become acquainted with the faculty's program prior to applying.
What it does
FACULTY FACTSHEET Detailed and fun information about each faculty QUIZ To represent the identity of individual faculty FACULTY RECOMMENDATION Suggesting related faculties to explore
How we built it
We create the python code in google colab. The code imports data, prepares data then process with k-mean method and 10 iteration. Moreover, it also calculate silhouette score to help further analysis. After clustering, we bring the data to present in tableau for recommend faculty analysis.
Challenges we ran into
The analysis need to use various study field data. So the analysis can provide various recommend faculty.
Accomplishments that we're proud of
- we can found the data source of exam score in variety subjects(9subjects) individually
- we can select the best number of cluster with silhouette score.
What we learned
Clustering is capable of grouping only a few groups. Therefore, not suitable for categorizing faculties. Solution : recommend faculties following the criterion instead of recommend specific faculty.
What's next for Let's learn
To improve the project
- if the data source have more variety of study field, the recommend faculty can be more specific.
- the classification method may be one of the good method to apply with this project. We can try using classification method in the future.
Log in or sign up for Devpost to join the conversation.