Skip to content

Latest commit

 

History

History
53 lines (37 loc) · 2.05 KB

slide.md

File metadata and controls

53 lines (37 loc) · 2.05 KB
Raw

Data Science Captone: Yelp Data Analysis

author: Jaehyun Shin date: 21 Nov 2015

The purpose of this Study

  • Find the correlation between a variety of attribution and rating including additional service which business provides.

  • Build rating prediction model based on attribute data.

Business Attributes And Category

  • Category (783 categories exist.)
[1] "Doctors"          "Health & Medical" "Nightlife"       
[4] "Active Life"      "Mini Golf"        "Golf"            
[7] "Shopping"         "Home Services"
  • Attributes (78 attributes exist.)
[1] "By.Appointment.Only"  "Happy.Hour"           "Accepts.Credit.Cards"
[4] "Good.For.Groups"      "Outdoor.Seating"      "Price.Range"         
[7] "Good.for.Kids"        "Alcohol"

Methodology

Model an prediction algorithm through the relevant data. Since the relevant data is to measure the correlation between attributes and stars, "classification modeling" will be used, and "rpart", a type of the algorithm, will be adopted.

      RMSE   Rsquared 
0.70710678 0.03560378

The error level is comparatively high when using postResample to measure. But while considering the fact that the relevant value is rating, the difference between the estimation and the actual measured value is examined.

Final Analysis

plot of chunk unnamed-chunk-5

Cases in which the error range of the ratings is below 0.5 is 72.1830986% of all data.

As shown on the graph, the difference between the estimation and actual measured value is mostly concentrated between -0.5 ~ 0.5, and since the median value and average value are very close to 0, the data is reliable.