Aerofly

• 

  Sign up

• 

  Prediction page

• 

  landing -loading page

Inspiration

Many people travel via airlines everyday , and while 80% of the flights are on time , the left 20% of them are delayed .This leads to a lot of time waste.Thus , I thought of coming up with a flight delay prediction model that can be of great help to both the passangers as well as the airlines.

What it does

It predicts the probability of the airplane being delayed on the basis of time of flight, length of flight , airline, airport locations.It can be of use to passangers who want to check if a particular flight will be delayed or not prior to booking the flight ,and also help airlines ,by giving them a signal to check for the ontime transportation if the model shows that a particular flight has the probability of getting delayed.

How we built it

I first started with building the frontend for the project using react, while training the data on the model in parallel to come up with the best model that gives the maximum accuracy.The model was then built on random forest model. After that I built the backend to get the data from the users (Signup and Signin)using Mongodb and in the very end , i found a way to connect the nodejs javascript to the python model.In the very end , i wrote a detailed readme file for the project.

Challenges we ran into

->Connecting the nodejs to the python model -> Finding a way to reduce the size of the model as it was quite heavy

Accomplishments that we're proud of

->This is the first time that I connected a javascript framework to python , so I am very proud of it , becuse it is recommended that if you have backend in python ,have the frontend also in python, still i figured out a way to solve the puzzle .

What we learned:

Quite a few things actually => ->We can build an entire project in such a less amount of time if we work with focus ->Their are many problems that can be solved with tech

What's next for Aerofly

->Incorporating weather data features to increase the model accuracy ->Introducing more personalised environment by translating the web app in the preferred language of the user ->Making the model fast ->Reducing the model size

Built With

• express.js
• machine-learning
• mongodb
• node.js
• python
• react