Skip to content

Latest commit

 

History

History
18 lines (11 loc) · 880 Bytes

README.md

File metadata and controls

18 lines (11 loc) · 880 Bytes

Starbucks-Location-Predictor

#This will predict if a Starbucks store will be successful in a particular location or not.

  • This code uses the data from opendata.socrata.com to get the store number, latitude and longitude.

  • Using this lat and long values, the FetchJSONObject.java code retrieves the attributes around each starbuck location. This is done using google places API.

  • The google places api url sample is https://maps.googleapis.com/maps/api/place/nearbysearch/json?location=43.031873,-71.073203&radius=500&key=

  • This .xlsx file is converted into .arff to give it to the Weka API

  • Using Weka API, classification is performed to classify data into two sets, one is the store which is in a big city and other is the store which is in rural area.

  • Naive Bayes is used to predict what is the probability of a store being successful in a particular location