The goal of the project was to develop a Python application that people could use and potentially monetize. We decided to develop an application that would provide awareness and incentives to users and help them decide among various modes of transportation in Manhattan - walking, biking, or taking a taxi.
The application takes in the time, day, origin, and destination and estimates the trip time and fare of taking a taxi, as well as the benefits (calories burned) of walking or biking. In addition, it would help bring foot traffic to businesses along the route by advertising and displaying coupons to users. Users may be more willing to purchase healthy foods and coupons for activities knowing that they saved money by not taking a taxi.
The application uses the random forest algorithm in order to predict the time of taking a taxi and takes that input in order to predict the fare cost. Although we had to scale down the amount of data used for the random forest model significantly, it was still able to perform fairly well and was often around 1-2 minutes different from the estimated time provided by Google Maps (which gets a lot more data). The walking and biking formulas need further development to provide more accurate trip duration and calories burned.