API tools faq
paste
Advertisement
Nenkham

Untitled

Nenkham
Jun 5th, 2019
99
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
R 8.98 KB | None | 0 0
raw download clone embed print report
  1.  
  2.  
  3.  
  4. # AS NOVAS FEATURES...
  5.  
  6.  
  7.  
  8.  
  9. TREINO <- features_treino
  10. VALID <- features_validacao
  11. TESTE <- features_teste
  12.  
  13.  
  14.  
  15.  
  16. ####TRANSFORMACOES PARA DECISION TREE...
  17.  
  18. # TREINO
  19.  
  20. TREINO <- TREINO[,-c(1)] # retiramos cartao
  21. TREINO <- TREINO[,-c(3)] # retiramos district
  22.  
  23.  
  24. # VALID
  25. VALID <- VALID[,-c(1)] # retiramos cartao
  26. VALID <- VALID[,-c(3)] # retiramos district
  27.  
  28.  
  29.  
  30. ####TRANSFORMACOES PARA RANDOM FOREST...
  31.  
  32. # TREINO
  33. f_TREINO <- TREINO
  34. # ja retiramos cartao e districs aqui!!!
  35. f_TREINO$Comprou <- as.factor(f_TREINO$Comprou) # transformar a label em factor para dar como input na randomForest(classificação)
  36.  
  37. f_TREINO[,3] <- ifelse(f_TREINO[,3] == "NORTE  ", 1, ifelse(f_TREINO[,3] == "SUL   ", 2, ifelse(f_TREINO[,3] == "CENTRO ", 3, ifelse(f_TREINO[,3] == "MADEIRA", 4, ifelse(f_TREINO[,3] == "AÇORES ", 5,6)))))
  38. f_TREINO[,4] <- ifelse(f_TREINO[,4] == "F", 1, ifelse(f_TREINO[,4] == "M", 2, ifelse(f_TREINO[,4] == "A", 3, 4)))
  39.  
  40.  
  41. # VALID
  42. f_VALID <- VALID
  43. # ja retiramos cartao e districs aqui!!!
  44. # nao queremos transformar a Label de validacao para ja
  45. f_VALID[,3] <- ifelse(f_VALID[,3] == "NORTE  ", 1, ifelse(f_VALID[,3] == "SUL   ", 2, ifelse(f_VALID[,3] == "CENTRO ", 3, ifelse(f_VALID[,3] == "MADEIRA", 4, ifelse(f_VALID[,3] == "AÇORES ", 5,6)))))
  46. f_VALID[,4] <- ifelse(f_VALID[,4] == "F", 1, ifelse(f_VALID[,4] == "M", 2, ifelse(f_VALID[,4] == "A", 3, 4)))
  47.                        
  48. # --------------------------------------------------------------------------------------------------------------------------#
  49. # FORM
  50.  
  51.  
  52.  
  53. features <- c('age', 'region', 'genero', 'total_compras3', 'total_AL3', 'total_semana3', 'media_semana3', 'total_FDS3', 'media_FDS3', 'total_online3', 'total_AUT3','total_bazar3', 'total_bio3', 'total_desporto3', 'total_lar3', 'total_pets3', 'total_restau3', 'total_roupa3', 'total_saude3', 'total_transp3', 'total_viagens3', 'total_compras6', 'total_AL6', 'total_semana6', 'media_semana6', 'total_FDS6', 'media_FDS6', 'total_online6', 'qte_online6', 'total_AUT6', 'total_bazar6', 'total_bio6', 'total_desporto6', 'total_lar6', 'total_pets6', 'total_restau6', 'total_roupa6', 'total_saude6', 'total_transp6', 'total_viagens6','total_compras12', 'total_AL12', 'total_semana12', 'media_semana12', 'total_FDS12', 'media_FDS12', 'total_online12', 'qte_online12', 'total_AUT12', 'total_bazar12', 'total_bio12', 'total_desporto12', 'total_lar12', 'total_pets12', 'total_restau12', 'total_roupa12', 'total_saude12', 'total_transp12', 'total_viagens12')
  54. label <- c('Comprou')
  55.  
  56.  
  57. form <- as.formula(paste(paste(label, " ~ "), paste(features, collapse = "+")))
  58. # --->PRECISION E RECALL
  59. metrics <- function(cm) {
  60.  
  61.   tp <- cm['table']$table[4]
  62.   tn <- cm['table']$table[1]
  63.   fp <- cm['table']$table[2]
  64.   fn <- cm['table']$table[3]
  65.  
  66.   precision <- tp / (tp + fp)
  67.   recall <- tp / (tp + fn)
  68.  
  69.   values <- c(precision, recall)
  70.  
  71.   return(values)
  72.  
  73. }        
  74.  
  75.  
  76.  
  77.  
  78.  
  79. # --------------------------------------------------------------------------------------------------------------------------#
  80. # ---------- MODELOS --------- #
  81. # --------------------------------------------------------------------------------------------------------------------------#
  82.  
  83. # --------------------------- #
  84. # ------ DECISION TREE ------ #
  85. # --------------------------- #
  86.  
  87. # -------> decision tree modelo Default
  88. tree_model_0 <- rpart(form, TREINO)
  89. rpart.plot(tree_model_0)
  90. # ---> ROC
  91. preds_tree_model_0 <- predict(tree_model_0, VALID)
  92.  
  93. threshold_preds <- (preds_tree_model_0 > 0.5) * 1
  94.  
  95. roc_curve_tree_0 <- roc.curve(scores.class0 = threshold_preds, weights.class0 = VALID[,label], curve=TRUE)
  96.  
  97. plot(roc_curve_tree_0)
  98. # ---> matriz de confusao
  99. cm_tm_0 <- confusionMatrix(as.factor(threshold_preds), as.factor(VALID[,label]))
  100. # ---> precision e recall
  101. metrics(cm_tm_0)
  102. # --------------------------------------------------------------------------------------------------------------------------#
  103.  
  104.  
  105. # --------> tree_model 1
  106. tree_model_1 <- rpart(form, TREINO, control = rpart.control(minsplit = 30, cp = 0.015, maxdepth = 30))
  107. rpart.plot(tree_model_1)
  108. # ---> ROC
  109. preds_tree_model_1 <- predict(tree_model_1, VALID)
  110. threshold_preds_1 <- (preds_tree_model_1 > 0.2) * 1
  111. roc_curve_tree_1 <- roc.curve(scores.class0 = preds_tree_model_1, weights.class0 = VALID[,label], curve=TRUE)
  112. plot(roc_curve_tree_1)
  113. # ---> matriz de confusao
  114. cm_tm_1 <- confusionMatrix(as.factor(threshold_preds_1), as.factor(VALID[,label]))
  115. # ---> precision e recall
  116. metrics(cm_tm_1)
  117. # --------------------------------------------------------------------------------------------------------------------------#
  118.  
  119.  
  120.  
  121.  
  122. # --------------------------------------------------------------------------------------------------------------------------#
  123. # ------ RANDOM FOREST ------ #
  124. # --------------------------- #
  125.  
  126. # --------> randomForest modelo Default
  127. rf_model_0 <- randomForest(form, f_TREINO, type="prob")
  128. print(rf_model_0)
  129. # ---> ROC
  130. preds_rf_model_0 <- predict(rf_model_0, f_VALID, type="prob")
  131.  
  132. preds_rf_model_0  <- preds_rf_model_0[,-c(1)] #esta a dar-nos prob da class 0 e prob da class 1, vamos retirar a variavel 0
  133. threshold_preds_rf_0 <- (preds_rf_model_0 > 0.5) * 1
  134.  
  135. roc_curve_rf_0 <- roc.curve(scores.class0 = threshold_preds_rf_0, weights.class0 = f_VALID[,label], curve=TRUE)
  136. plot(roc_curve_rf_0)
  137. # ---> matriz de confusao
  138. cm_rf_0 <- confusionMatrix(as.factor(threshold_preds_rf_0), as.factor(f_VALID[,label]))
  139. # ---> precision e recall
  140. metrics(cm_rf_0)
  141. # --------------------------------------------------------------------------------------------------------------------------#
  142.  
  143.  
  144. # --------> randomForest_1
  145. rf_model_1 <- randomForest(form, f_TREINO, type="prob",  ntree=30000, mtry=4) #AUC=0.670   Pre=0.675   Recall=0.432
  146. print(rf_model_1)
  147. # ---> ROC
  148. preds_rf_model_1 <- predict(rf_model_1, f_VALID, type="prob")
  149.  
  150. preds_rf_model_1  <- preds_rf_model_1[,-c(1)] #esta a dar-nos prob da class 0 e prob da class 1, vamos retirar a variavel 0
  151. threshold_preds_rf_1 <- (preds_rf_model_1 > 0.5) * 1
  152.  
  153. roc_curve_rf_1 <- roc.curve(scores.class0 = threshold_preds_rf_1, weights.class0 = f_VALID[,label], curve=TRUE)
  154. plot(roc_curve_rf_1)
  155. # ---> matriz de confusao
  156. cm_rf_1 <- confusionMatrix(as.factor(threshold_preds_rf_1), as.factor(f_VALID[,label]))
  157. # ---> precision e recall
  158. metrics(cm_rf_1)
  159.  
  160.  
  161.  
  162.  
  163. # --------> randomForest_DMwR2
  164.  
  165. potato <- rpartXse(form, TREINO)
  166. prp(potato,type=4,extra=101)
  167. potato_preds <- predict(potato,VALID)
  168.  
  169. potato_mae <- mean(abs(potato_preds-VALID[,label]))
  170.  
  171. potato_mae
  172.  
  173. cor_potato <- cor(potato_preds,VALID[,label])
  174.  
  175. cor_potato
  176.  
  177. potato_res <- performanceEstimation(
  178.                 PredTask(form, TREINO),
  179.                 Workflow(
  180.                           "standardWF",
  181.                           learner="rpartXse",
  182.                           learner.pars=list(se=c(0,0.5,1)),
  183.                           predictor.pars=list(type="class")
  184.                         ),
  185.                
  186.                 EstimationTask(metrics="err",method=CV(nReps=3,nFolds=10)))
  187.  
  188. summary(potato_res)
  189. plot(potato_res)
  190.  
  191.  
  192.  
  193.  
  194.  
  195. # --------------------------------------------------------------------------------------------------------------------------#
  196.  
  197.  
  198.  
  199. # --------------------------------------------------------------------------------------------------------------------------#
  200. # ----------- SVM (C-classification) ----------- #
  201. # ---------------------------------------------- #
  202.  
  203. # --------> SVM modelo baseline # RADIAL # ----------- #
  204. svm_model_R0 <- svm(form, TREINO, type="C-classification", kernel="radial")
  205. print(svm_model_R0)
  206. # ---> ROC
  207. preds_svm_model_R0 <- predict(svm_model_R0, VALID)
  208. roc_curve_svm_R0 <- roc.curve(scores.class0 = preds_svm_model_R0, weights.class0 = VALID[,label], curve=TRUE)
  209. plot(roc_curve_svm_R0)
  210. # ---> matriz de confusao
  211. cm_svm_R0 <- confusionMatrix(as.factor(preds_svm_model_R0), as.factor(VALID[,label]))
  212. # ---> precision e recall
  213. metrics(cm_svm_R0)
  214. # --------------------------------------------------------------------------------------------------------------------------#
  215.  
  216.  
  217. # ---> SVM 1 (radial)
  218. svm_model_R1 <- svm(form, TREINO, type="C-classification", kernel="radial", cost=50, gamma=1)
  219. print(svm_model_R1)
  220. # ---> ROC
  221. preds_svm_model_R1 <- predict(svm_model_R1, VALID)
  222. roc_curve_svm_R1 <- roc.curve(scores.class0 = preds_svm_model_R1, weights.class0 = VALID[,label], curve=TRUE)
  223. plot(roc_curve_svm_R1)
  224. # ---> matriz de confusao
  225. cm_svm_R1 <- confusionMatrix(as.factor(preds_svm_model_R1), as.factor(VALID[,label]))
  226. # ---> precision e recall
  227. metrics(cm_svm_R1)
  228. # --------------------------------------------------------------------------------------------------------------------------#
  229.  
  230.  
  231.  
  232. # --------------------------------------------------------------------------------------------------------------------------#
  233. # ------------------------------------------------> END <-------------------------------------------------------------------#
  234. # --------------------------------------------------------------------------------------------------------------------------#
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement
Public Pastes
Advertisement
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!