Synopsis

The project is about HAR - Human Activity Recognition , where 6 participants were asked to perform barbell lifts correctly and incorrectly in 5 different ways and the relevant data were collected using the devices like Jawbone Up, Nike FuelBand, and Fitbit which were worn by the paricipants.

The participant regularly quantify how much of a particular activity they do, but they rarely quantify how well they do it.

The dataset contains data on 5 classes (sitting-down, standing-up, standing, walking, and sitting) collected on 8 hours of activities through those activity monitor devices.

The aim is to build a machine learning algorithm to predict activity quality from activity sensor monitors and train a model based on the data of various sensor values, which could later be used to predict the Classe variable, that is the manner in which the participants of HAR did the exercise.

Initial Setup

Enivironment

Hardware :
            Macbook Pro with OSX Yosemite 10.10.4
Software :
            RStudio 
            
            GitHub DeskTop / Web Version

Data Processing

Setting up working directory:

setwd("~/Desktop/Machine Learning/Coursera-PML-Project")

The required R Packeges were installed :

library(caret)
## Loading required package: lattice
## Loading required package: ggplot2
library(rpart)
library(rpart.plot)
library(randomForest)
## randomForest 4.6-10
## Type rfNews() to see new features/changes/bug fixes.
library(corrplot)

Download the Data :

        The training and test data were downloaded as instructed.
        Trainig Data : pml-training.csv 
        Test data    : pml-testing.csv
        
trainUrl <-"https://d396qusza40orc.cloudfront.net/predmachlearn/pml-training.csv"
testUrl <- "https://d396qusza40orc.cloudfront.net/predmachlearn/pml-testing.csv"
trainFile <- "./data/pml-training.csv"
testFile  <- "./data/pml-testing.csv"
if (!file.exists("./data")) {
  dir.create("./data")
}
if (!file.exists(trainFile)) {
  download.file(trainUrl, destfile=trainFile, method="curl")
}
if (!file.exists(testFile)) {
  download.file(testUrl, destfile=testFile, method="curl")
}

Read the data into Dataframes and check for the content:

trainRaw <- read.csv("./data/pml-training.csv")
testRaw <- read.csv("./data/pml-testing.csv")
dim(trainRaw)
## [1] 19622   160
dim(testRaw)
## [1]  20 160

Cleaning the Data :

On inspecting both traing and test the dataset, some missing data and unwated data were found which were to be cleaned.

sum(complete.cases(trainRaw))
## [1] 406

The columns containing NA were replaced with 0:

trainRaw <- trainRaw[, colSums(is.na(trainRaw)) == 0] 
testRaw <- testRaw[, colSums(is.na(testRaw)) == 0]

Few columns where those data do not have any meaning with the measurements data were removed.

classe <- trainRaw$classe
trainRemove <- grepl("^X|timestamp|window", names(trainRaw))
trainRaw <- trainRaw[, !trainRemove]
trainCleaned <- trainRaw[, sapply(trainRaw, is.numeric)]
trainCleaned$classe <- classe
testRemove <- grepl("^X|timestamp|window", names(testRaw))
testRaw <- testRaw[, !testRemove]
testCleaned <- testRaw[, sapply(testRaw, is.numeric)]

The cleaned training data set now contains 19622 observations and 53 variables, while the testing data set contains 20 observations and 53 variables. The “classe” variable is still in the cleaned training set.

Slicing the data:

Then, we can split the cleaned training set into a pure training data set (70%) and a validation data set (30%). We will use the validation data set to conduct cross validation in future steps.

set.seed(22519) # For reproducibile purpose
inTrain <- createDataPartition(trainCleaned$classe, p=0.70, list=F)
trainData <- trainCleaned[inTrain, ]
testData <- trainCleaned[-inTrain, ]

Data Modeling:

We fit a predictive model for activity recognition using Random Forest algorithm because it automatically selects important variables and is robust to correlated covariates & outliers in general. We will use 5-fold cross validation when applying the algorithm.

controlRf <- trainControl(method="cv", 5)
modelRf <- train(classe ~ ., data=trainData, method="rf", trControl=controlRf, ntree=250)
modelRf
## 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: 10989, 10989, 10991, 10990, 10989 
## 
## Resampling results across tuning parameters:
## 
##   mtry  Accuracy   Kappa      Accuracy SD  Kappa SD   
##    2    0.9910462  0.9886729  0.001301269  0.001647776
##   27    0.9914102  0.9891334  0.001717547  0.002174708
##   52    0.9850037  0.9810264  0.002718384  0.003439965
## 
## Accuracy was used to select the optimal model using  the largest value.
## The final value used for the model was mtry = 27.

The performance of the model on the validation data set is estimated.

predictRf <- predict(modelRf, testData)
confusionMatrix(testData$classe, predictRf)
## Confusion Matrix and Statistics
## 
##           Reference
## Prediction    A    B    C    D    E
##          A 1672    1    0    0    1
##          B    7 1128    4    0    0
##          C    0    0 1021    5    0
##          D    0    0   14  949    1
##          E    0    0    1    7 1074
## 
## Overall Statistics
##                                          
##                Accuracy : 0.993          
##                  95% CI : (0.9906, 0.995)
##     No Information Rate : 0.2853         
##     P-Value [Acc > NIR] : < 2.2e-16      
##                                          
##                   Kappa : 0.9912         
##  Mcnemar's Test P-Value : NA             
## 
## Statistics by Class:
## 
##                      Class: A Class: B Class: C Class: D Class: E
## Sensitivity            0.9958   0.9991   0.9817   0.9875   0.9981
## Specificity            0.9995   0.9977   0.9990   0.9970   0.9983
## Pos Pred Value         0.9988   0.9903   0.9951   0.9844   0.9926
## Neg Pred Value         0.9983   0.9998   0.9961   0.9976   0.9996
## Prevalence             0.2853   0.1918   0.1767   0.1633   0.1828
## Detection Rate         0.2841   0.1917   0.1735   0.1613   0.1825
## Detection Prevalence   0.2845   0.1935   0.1743   0.1638   0.1839
## Balanced Accuracy      0.9977   0.9984   0.9903   0.9922   0.9982
accuracy <- postResample(predictRf, testData$classe)
accuracy
##  Accuracy     Kappa 
## 0.9930331 0.9911872
outSampErr <- 1 - as.numeric(confusionMatrix(testData$classe, predictRf)$overall[1])
outSampErr
## [1] 0.006966865

The accuracy of the model is estimated as 99.30% and out of sample error of 0.69 % is estimated.

Predicting for Test Data Set:

Finally we apply the model to the original testing data set downloaded from the data source. We remove the problem_id column first.

result <- predict(modelRf, testCleaned[, -length(names(testCleaned))])
result
##  [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

Conclusion

It is found that the accuracy of the model is 99.30% and out of sample error is 0.69% .

Plots are available in the folder ‘writeup_files/figure-html’ in th GitHub Repository and appended in Appendix.

I acknowledge my thanks to the Coursera team and course participants for giving this opporunity to enrich my knowledge by doing this interesting project.

I also thank HAR data provider for this excercise.

Appendix

Plot 1: Decision Tree Visualization
treeModel <- rpart(classe ~ ., data=trainData, method="class")
prp(treeModel)
Plot 2 :Correlation Matrix Visualization
corrPlot <- cor(trainData[, -length(names(trainData))])
corrplot(corrPlot, method="color")