In armandvalsesia/FoodC5: C5.0 classifier for Food Items Mapping

library(C50)
library(pander)
library(FoodC5)

Retrieve fuzzy scores and predict mapping using a pre-trained model

Retrieve pre-computed fuzzy scores for English food names and use a pre-trained C5.0 model to classify results as good or bad matches. Fuzzy matching scores can be computed using the Foodmapping R package.

data(food_score)
preds <- c5_one_language(food_score,  'ENG_SCORE', 'EDIFF')
pander(head(preds), split.table = Inf )

The same approach can also be made, using a comparison of the original food name.

preds <- c5_one_language(food_score,  'ORI_SCORE', 'EDIFF', c5_object = read_c50_ori() ) 
pander(head(preds), split.table = Inf )

Explore the C5.0 tree

Summary of the defaut English Based C5.O tree

eng_tree <- read_c50_eng()
summary(eng_tree)

armandvalsesia/FoodC5 documentation built on May 6, 2019, 12:23 p.m.