Schedule Generator by OurSQL

Directory guide:

• db contains necessary instructions for MySQL server initialization and connection on Amazon RDS, as well as other SQL scripts
• server contains Flask server backend instructions, Amazon EC2 server setup instructions, and code for fetching/transferring data
• client contains JS user-facing code
• model contains the neural network that powers schedule generation and recommendation

Local dependencies:

• A local SSH client is necessary to connect to AWS
• Installation of Neo4J Desktop is recommended
• Python3 (conda distribution preferred):
  • conda install numpy pandas pathlib tqdm neo4j pymysql sqlalchemy flask flask-cors -y
  • pip install neo4j python-dotenv
  • conda install pytorch cpuonly -c pytorch -y

Setup Script

To install the module run python setup.py install. To clean the module from the directory run bash clean_install.sh

Advanced Function 1

The advanced function is used to generate an optimal schedule. Scores are calculated based off of this formula:

There are multiple factors that go into this:

• AVG_GPA is the average GPA of all the classes in a candidate schedule
• AVG_User_Rating is the average of average ratings across all user-provided ratings for the classes in the schedule
• AVG_NN_Score is the average NN rating. This rating takes into account temporal trends
• num_interest_courses and num_courses correspond to how many courses align with a student's stream and the number of courses in the schedule
• num_CH and num_CH_desired correspond to the total credit-hours of the schedule, and the amount the user wants to limit to.