Skip to content

Latest commit

 

History

History
172 lines (136 loc) · 4.92 KB

PredictionAssignment.md

File metadata and controls

172 lines (136 loc) · 4.92 KB
Raw
title author date output
Prediction Assignment
Babita
4 September 2020
html_document
keep_md
true

Background

Using devices such as Jawbone Up, Nike FuelBand, and Fitbit it is now possible to collect a large amount of data about personal activity relatively inexpensively. These type of devices are part of the quantified self movement – a group of enthusiasts who take measurements about themselves regularly to improve their health, to find patterns in their behavior, or because they are tech geeks. One thing that people regularly do is quantify how much of a particular activity they do, but they rarely quantify how well they do it. In this project, your goal will be to use data from accelerometers on the belt, forearm, arm, and dumbell of 6 participants. They were asked to perform barbell lifts correctly and incorrectly in 5 different ways.

library(RCurl)
library(caret)

Downloading Data

setwd("F:\\Work\\Techwalnut\\DS And ML\\Courses\\Coursera\\DS Specialization _JHU\\Course8_Practical_ML\\Week4\\Peer Assignment\\DS_C8_PredictionAssignment")

if (!file.exists("./pml-training.csv")) {
  url.training <- "https://d396qusza40orc.cloudfront.net/predmachlearn/pml-training.csv"
  download.file(url.training, destfile = "./pml-training.csv")
}

if (!file.exists("./pml-testing.csv")) {
  url.testing <- "https://d396qusza40orc.cloudfront.net/predmachlearn/pml-testing.csv"
  download.file(url.testing, destfile = "./pml-testing.csv")
}

Reading Data

trainData<- read.csv("./pml-training.csv")
testData<- read.csv("./pml-testing.csv")

Data Preprocessing

Removing N/As and unuseful columns

trainData <- trainData[, colSums(is.na(trainData)) == 0] 
testData <- testData[, colSums(is.na(testData)) == 0] 
classe <- trainData$classe
trainR <- grepl("^X|timestamp|window", names(trainData))
trainData <- trainData[, !trainR]
trainM <- trainData[, sapply(trainData, is.numeric)]
trainM$classe <- classe
testR <- grepl("^X|timestamp|window", names(testData))
testData<- testData[, !testR]
testM <- testData[, sapply(testData, is.numeric)]

Spliting Data into train and test

set.seed(12345) 
inTrain <- createDataPartition(trainM$classe, p=0.70, list=F)
train_data <- trainM[inTrain, ]
test_data <- trainM[-inTrain, ]

Random Forest Model

setting <- trainControl(method="cv", 5)
RandomForest <- train(classe ~ ., data=train_data, method="rf", trControl=setting,
                      ntree = 100)
RandomForest
## Random Forest 
## 
## 13737 samples
##    52 predictor
##     5 classes: 'A', 'B', 'C', 'D', 'E' 
## 
## No pre-processing
## Resampling: Cross-Validated (5 fold) 
## Summary of sample sizes: 10987, 10990, 10990, 10991, 10990 
## Resampling results across tuning parameters:
## 
##   mtry  Accuracy   Kappa    
##    2    0.9912639  0.9889488
##   27    0.9895169  0.9867389
##   52    0.9847859  0.9807534
## 
## Accuracy was used to select the optimal model using the largest value.
## The final value used for the model was mtry = 2.

Accuracy and out-of-sample error

predict_RandomForest <- predict(RandomForest, test_data)
confusionMatrix(test_data$classe, predict_RandomForest)
## Confusion Matrix and Statistics
## 
##           Reference
## Prediction    A    B    C    D    E
##          A 1673    1    0    0    0
##          B    5 1133    1    0    0
##          C    0    4 1022    0    0
##          D    0    0   24  938    2
##          E    0    0    0    1 1081
## 
## Overall Statistics
##                                           
##                Accuracy : 0.9935          
##                  95% CI : (0.9911, 0.9954)
##     No Information Rate : 0.2851          
##     P-Value [Acc > NIR] : < 2.2e-16       
##                                           
##                   Kappa : 0.9918          
##                                           
##  Mcnemar's Test P-Value : NA              
## 
## Statistics by Class:
## 
##                      Class: A Class: B Class: C Class: D Class: E
## Sensitivity            0.9970   0.9956   0.9761   0.9989   0.9982
## Specificity            0.9998   0.9987   0.9992   0.9947   0.9998
## Pos Pred Value         0.9994   0.9947   0.9961   0.9730   0.9991
## Neg Pred Value         0.9988   0.9989   0.9949   0.9998   0.9996
## Prevalence             0.2851   0.1934   0.1779   0.1596   0.1840
## Detection Rate         0.2843   0.1925   0.1737   0.1594   0.1837
## Detection Prevalence   0.2845   0.1935   0.1743   0.1638   0.1839
## Balanced Accuracy      0.9984   0.9972   0.9876   0.9968   0.9990

accuracy <- postResample(predict_RandomForest, test_data$classe)
error<-1 - as.numeric(confusionMatrix(test_data$classe, predict_RandomForest)$overall[1])
error
## [1] 0.006457094

Accuracy: 99.3542906%

out-of-sample error: 0.6457094%

Prediction on test data

predict(RandomForest, testM)
##  [1] 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