Birth-Weight-Prediction/Updated Analysis.R at master · Joseph1Otoo/Birth-Weight-Prediction · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
library(dplyr)

data=read.csv(file.choose(), header = T, na.strings = " ")

str(data)
summary(data)
View(data)

## REMOVE UNWANTED VARIABLE and MISSING DATA

data[,24]<-NULL

data <- filter(data, !(is.na(P_EDU)))
data <- filter(data, !(is.na(P_JBT)))
data <- filter(data, !(is.na(JBT)))
data <- filter(data, !(is.na(OCC)))
data <- filter(data, !(is.na(B_SIZE)))

## DROP 8 and 9 from B_SIZE

data <- droplevels(data %>% filter(B_SIZE != "8"))
data <- droplevels(data %>% filter(B_SIZE != "9"))

nrow(data)

#############CHANGING NUMERIC INTO FACTORS WITH THEIR LABELS#######

data$B_SIZE <- factor(data$B_SIZE,levels = c(1:5),labels = c("Very Large", "Larger than average","Average", "Smaller than average", "Very Small"))
data$PAR <- factor(data$PAR, levels = c(1:2), labels = c("Single", "Multiple"))
data$D_COMP <- factor(data$D_COMP, levels = c(1:3), labels = c("Born alive", "Born dead", "Lost pregnancy"))
data$AGE <- factor(data$AGE, levels = c(1:7), labels = c("15-19","20-24","25-29","30-34","35-39","40-44","45-49"))
data$RES <- factor(data$RES, levels = c(1:2), labels = c("Urban","Rural"))
data$C_RES <- factor(data$C_RES, levels = c(0:4), labels = c("Large city","City","Town","Countryside","Abroad"))
data$EDU <- factor(data$EDU, levels = c(0:3), labels = c("No education","Primary","Secondary","Higher"))
data$M_ST <-factor(data$M_ST, levels = c(0:5), labels = c("Never married","Married","Living together","Widowed","Divorced","Not living together"))

str(data)
summary(data)

## Categories of birth weight in percentages

T=table(data$B_SIZE)
T
T1=round(prop.table(T), digits = 2)
round(100*T1, digits = 0)

## Categories of delivery outcomes in percentages

T2=table(data$D_COMP)
T2
T3=round(prop.table(T2), digits = 2)
round(100*T3, digits = 0)

# Splitting the dataset into the Training set and Test set
library(caTools)

set.seed(123)
split = sample.split(data$B_SIZE, SplitRatio = 0.7)

training_set = subset(data, split == TRUE)
test_set = subset(data, split == FALSE)

table(training_set$B_SIZE)
str(training_set)

### NEW SVM APPROACH

library(e1071)
mymodel= svm(B_SIZE~ AGE+RES+EDU+PAR+C_RES+M_ST, data = training_set)
summary(mymodel)

attributes(mymodel)

##### MISCLASSIFICATION ERROR FOR TRAINING SET

pred1 = predict(mymodel, training_set)

tab1 = table(Predicted = pred1, Actual = training_set[,27])
tab1

mr1=1 - sum(diag(tab1))/sum(tab1)
round(mr1*100, digits = 2)

### In the plot when there more variables we use
## slice = list(v1=3,v2=4)
##It is worth noting that the plot is for continuos variables

# plot(mymodel, data, PAR~EDU, slice = list(AGE=3, RES=4))

## Confusion matrix and misclassification error

pred = predict(mymodel, test_set)

tab = table(Predicted = pred, Actual = test_set[,27])
tab

### MISCLASSIFICATION RATE

mr=1 - sum(diag(tab))/sum(tab)
round(mr*100, digits = 2)


### RANDOM FOREST

library(randomForest)

set.seed(222)

RMmodel = randomForest(B_SIZE ~ AGE+RES+EDU+PAR+C_RES+M_ST,data = training_set)

print(RMmodel)


#### MODEL AFTER TUNING PARAMETERS

RMmodel = randomForest(B_SIZE ~ AGE+RES+EDU+PAR+C_RES+M_ST,data = training_set,
                       ntree = 300, mtry = 10,
                       importance = TRUE,
                       proximity = TRUE)
print(RMmodel)


attributes(RMmodel)

RMmodel$terms

library(caret)

p1 = predict(RMmodel, training_set)
confusionMatrix(p1, training_set$B_SIZE)


p2 = predict(RMmodel, test_set)
confusionMatrix(p2, test_set$B_SIZE)

## ERROR RATE

plot(RMmodel)


## TUNE RF

t = tuneRF(training_set[,-27], training_set[,27],
           stepFactor = 0.5, plot = TRUE,
           ntreeTry = 300, trace = TRUE,
           improve = 0.05)

### NUMBER OF NODES FOR THE TREES

hist(treesize(RMmodel), main = "No. of nodes for trees",
     col = "blue")

## VARIABLE IMPORTANCE

varImpPlot(RMmodel, sort = TRUE,
           n.var = 6,
           main = "Top 6 variables of importance")
importance(RMmodel)
varUsed(RMmodel)


### PARTIAL DEPENDENCY PLOT

partialPlot(RMmodel, training_set, RES, "Average")


## EXTRACT SINGLE TREE

getTree(RMmodel, 1, labelVar = TRUE)


### MULTI- DIMENSIONAL  SCALING PLOT OF THE PROXIMITY MATRIX

MDSplot(RMmodel, training_set$B_SIZE)