PerfectDateSA

Inspiration

Flourishing tourism industry in San Antonio with a focus on enhancing the experience for tourists coming into San Antonio and for local business owners
Focus on the best use of San Antonio Data with the motive of the best times to travel to San Antonio ## What it does
Developed a predictive modeling algorithm using web scrapped data from various travel, tourism, and government websites to give the best possible combination of flight tickets, and hotel prices within the budget and date selected by the customer. -For the time interval selected by the customer, the model shows the events taking place in Six Flags and San Antonio thereby increasing the awareness among users and business owners -Gives the Busyness rating for Six Flag for the top 5 dates and budget suggested by the model, thereby enabling them to plan ahead ## How we built it
Used BeautifulSoap and Selenium for Web Scrapping the data from various websites
Used external APIs for getting flight-related data
Built and trained ARIMA machine learning model for the time-series data which was collected by the above methods and evaluated the model MSE, RMSE, MAE, and MAPE
Conducted Dickey-Fuller test and got statistics, view daily price changes, graphs for correlation
Built a front-end application on Streamlit for getting inputs from the user and displaying graphs, predictions, suggestions, events, and busyness ratings to the user ## Challenges we ran into
Getting the right data for solving the business question we had
Time constraints for building a back-end model, collecting data, and building a front-end ## Accomplishments that we're proud of
The San Antonio data that we scrapped out of the web gives an excellent suggestion on when to visit San Antonio and what would be the budget required for travel and stay. This suggestion has been tested on several aspects and we feel that it is pretty accurate. ## What we learned
The time to develop a machine learning model depends on the data that we have, hence data collection and processing is one such task on which attention is not given to, but is the most important task. ## What's next for PerfectDate SA -Personalized suggestions for users based on their linkings and onboarding business to the application for enhancing their business
a new UI, probably
bigger and better datasets to train and test the model