Machine Learning with R
2 likes774 views
This document downloads training and testing data files for a machine learning model. It then cleans the training data, removing columns with missing or non-numeric values. A random forest model is trained on 70% of the cleaned training data and predicts classes for the remaining 30%, achieving 99.4% accuracy based on a confusion matrix. The trained model is then used to predict classes for the cleaned testing data.
![dim(pmlTraining)
## [1] 19622 160
dim(pmlTesting)
## [1] 20 160](https://image.slidesharecdn.com/pt2cdwi0rgm8irnteh1e-signature-abf04e2c917e23a06995ba2bd3bba28948a03bcdebd5b3992b2cc0b0c410b603-poli-141008173638-conversion-gate02/85/Machine-Learning-with-R-6-320.jpg)
![ProcessData <- function (dataframe){
col <- vector(mode = "numeric")
for(i in 1:ncol(dataframe))
{
if(colnames(dataframe)[i]=='classe'){next;}
total <- length(dataframe[,i]);
NAs <- length(dataframe[is.na(dataframe[,i]),i]);
empty <- length(dataframe[dataframe[,i]=='',i])
notNumber <- length(dataframe[!is.numeric(dataframe[,i]),i]);
if(round(NAs/total)==1 | round(empty/total)==1 | round(notNumber/total)==1){
col <- rbind(col, i)
}
}
process <- dataframe[,as.numeric(col)*(-1)]
for(i in 1:(ncol(process)-1)){ process[,i] <- as.numeric(process[,i]); }
process[,c(1,2,3,4)*(-1)]
}](https://image.slidesharecdn.com/pt2cdwi0rgm8irnteh1e-signature-abf04e2c917e23a06995ba2bd3bba28948a03bcdebd5b3992b2cc0b0c410b603-poli-141008173638-conversion-gate02/85/Machine-Learning-with-R-10-320.jpg)
![cleanedData <- ProcessData(pmlTraining)
dim(cleanedData)
## [1] 19622 53](https://image.slidesharecdn.com/pt2cdwi0rgm8irnteh1e-signature-abf04e2c917e23a06995ba2bd3bba28948a03bcdebd5b3992b2cc0b0c410b603-poli-141008173638-conversion-gate02/85/Machine-Learning-with-R-11-320.jpg)
![library(caret)
inTrain <- createDataPartition(y = cleanedData$classe, p = 0.7, list = FALSE)
trainSet <- cleanedData[inTrain,]
validationSet <- cleanedData[-inTrain,]
dim(trainSet); dim(validationSet)
## [1] 13737 53
## [1] 5885 53](https://image.slidesharecdn.com/pt2cdwi0rgm8irnteh1e-signature-abf04e2c917e23a06995ba2bd3bba28948a03bcdebd5b3992b2cc0b0c410b603-poli-141008173638-conversion-gate02/85/Machine-Learning-with-R-14-320.jpg)
![predictResult <- predict(fit, validationSet[,-53])
confusionMatrix(validationSet[,53], predictResult)](https://image.slidesharecdn.com/pt2cdwi0rgm8irnteh1e-signature-abf04e2c917e23a06995ba2bd3bba28948a03bcdebd5b3992b2cc0b0c410b603-poli-141008173638-conversion-gate02/85/Machine-Learning-with-R-17-320.jpg)
![## Confusion Matrix and Statistics
##
## Reference
## Prediction A B C D E
## A 1673 1 0 0 0
## B 8 1128 3 0 0
## C 0 8 1014 4 0
## D 0 0 8 956 0
## E 0 0 0 1 1081
##
## Overall Statistics
##
## Accuracy : 0.994
## 95% CI : (0.992, 0.996)
## No Information Rate : 0.286
## P-Value [Acc > NIR] : <2e-16
##
## Kappa : 0.993
## Mcnemar's Test P-Value : NA
##
## Statistics by Class:
##
## Class: A Class: B Class: C Class: D Class: E
## Sensitivity 0.995 0.992 0.989 0.995 1.000
## Specificity 1.000 0.998 0.998 0.998 1.000
## Pos Pred Value 0.999 0.990 0.988 0.992 0.999
## Neg Pred Value 0.998 0.998 0.998 0.999 1.000
## Prevalence 0.286 0.193 0.174 0.163 0.184
## Detection Rate 0.284 0.192 0.172 0.162 0.184
## Detection Prevalence 0.284 0.194 0.174 0.164 0.184
## Balanced Accuracy 0.998 0.995 0.993 0.997 1.000](https://image.slidesharecdn.com/pt2cdwi0rgm8irnteh1e-signature-abf04e2c917e23a06995ba2bd3bba28948a03bcdebd5b3992b2cc0b0c410b603-poli-141008173638-conversion-gate02/85/Machine-Learning-with-R-18-320.jpg)
Machine Learning with R
- 5. download.file(url = "https://d396qusza40orc.cloudfront.net/predmachlearn/pml-training.csv", destfile = "pml-training.csv") download.file(url = "https://d396qusza40orc.cloudfront.net/predmachlearn/pml-testing.csv", destfile = "pml-testing.csv") pmlTraining <- read.csv("pml-training.csv") pmlTesting <- read.csv("pml-testing.csv")
- 6. dim(pmlTraining) ## [1] 19622 160 dim(pmlTesting) ## [1] 20 160
- 10. ProcessData <- function (dataframe){ col <- vector(mode = "numeric") for(i in 1:ncol(dataframe)) { if(colnames(dataframe)[i]=='classe'){next;} total <- length(dataframe[,i]); NAs <- length(dataframe[is.na(dataframe[,i]),i]); empty <- length(dataframe[dataframe[,i]=='',i]) notNumber <- length(dataframe[!is.numeric(dataframe[,i]),i]); if(round(NAs/total)==1 | round(empty/total)==1 | round(notNumber/total)==1){ col <- rbind(col, i) } } process <- dataframe[,as.numeric(col)*(-1)] for(i in 1:(ncol(process)-1)){ process[,i] <- as.numeric(process[,i]); } process[,c(1,2,3,4)*(-1)] }
- 11. cleanedData <- ProcessData(pmlTraining) dim(cleanedData) ## [1] 19622 53
- 14. library(caret) inTrain <- createDataPartition(y = cleanedData$classe, p = 0.7, list = FALSE) trainSet <- cleanedData[inTrain,] validationSet <- cleanedData[-inTrain,] dim(trainSet); dim(validationSet) ## [1] 13737 53 ## [1] 5885 53
- 16. library(caret) fit <- train(classe ~ ., data = trainSet, method = 'rf') library(randomForest) fit <- randomForest(classe ~ ., data = trainSet)
- 17. predictResult <- predict(fit, validationSet[,-53]) confusionMatrix(validationSet[,53], predictResult)
- 18. ## Confusion Matrix and Statistics ## ## Reference ## Prediction A B C D E ## A 1673 1 0 0 0 ## B 8 1128 3 0 0 ## C 0 8 1014 4 0 ## D 0 0 8 956 0 ## E 0 0 0 1 1081 ## ## Overall Statistics ## ## Accuracy : 0.994 ## 95% CI : (0.992, 0.996) ## No Information Rate : 0.286 ## P-Value [Acc > NIR] : <2e-16 ## ## Kappa : 0.993 ## Mcnemar's Test P-Value : NA ## ## Statistics by Class: ## ## Class: A Class: B Class: C Class: D Class: E ## Sensitivity 0.995 0.992 0.989 0.995 1.000 ## Specificity 1.000 0.998 0.998 0.998 1.000 ## Pos Pred Value 0.999 0.990 0.988 0.992 0.999 ## Neg Pred Value 0.998 0.998 0.998 0.999 1.000 ## Prevalence 0.286 0.193 0.174 0.163 0.184 ## Detection Rate 0.284 0.192 0.172 0.162 0.184 ## Detection Prevalence 0.284 0.194 0.174 0.164 0.184 ## Balanced Accuracy 0.998 0.995 0.993 0.997 1.000
- 19. cleanedTest <- ProcessData(pmlTesting) predict(fit, cleanedTest) ## 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 ## B A B A A E D B A A B C B A E E A B B B ## Levels: A B C D E