KNN.Rmd

---
title: "KNN"
author: "Mathew Yang"
output: html_document
---

#K-Nearest Neighbors

I will be using KNN to classify the tumor cells.
```{r}
library(class)
wdbc = read.csv("data/wdbc.data", header= FALSE)
```

We know that the cell area skews our dataset so we will scale it here to minimize biases.
```{r}
#Scales data by subtracting mean, then dividing by standard deviation.
scaled.wdbc = scale(wdbc[,3:32])
```

I will split the 569 observations into a training and testing set.
```{r}
#Matrix of predictors associated with training data
train.X = scaled.wdbc[-(1:200),]

#Matrix of predictors with test
test.X = scaled.wdbc[1:200,]

#Vector of class labels for training/testing
train.Y = wdbc$V2[-(1:200)]
test.Y = wdbc$V2[1:200]
```

```{r}
#Run the KNN model
set.seed(1)
knn.pred = knn(train.X, test.X, train.Y, k=1)

pred.result = mean(test.Y!=knn.pred) #Our error rate
print(paste("The error rate of the KNN model is ", pred.result))
```

KNN resulted in a 95.5% accuracy rate.