README.md

k-nearest neighbors (KNN)

Load Dataset

Load iris.csv to two separate numpy arrays X, y

Load features into X [sepal.length, sepal.width, petal.length, petal.width]

Load labels into y [0,1]

Shuffling

Randomize X and corresponding y simultaneously

Train Test Split

X_train = first 80% of X y_train = first 80% of y X_test = rest X y_test = rest y

Algorithm [Test Prediction]

k = 5
X_train = (M, N) # N columns with M rows [for this case N should be 4]

y_train = (M, 1) # 1 columns with M rows

X_test = (M’, N) # N columns with M’ rows [for this case N should be 4]

y_test = (M’, 1) # 1 columns with M’ rows

y_test_predicted = new numpy array of size M’
for i in range(len(X_test)):
	x_test = X_test[i]
	D = new numpy array of size M
	D = Calculate euclidean distances between x_test and X_train
	min_dist_indices  = find k indices in D where values are minimum
	y_neighbor = y_train[ min_dist_indices ]
	y_test_predicted[ i ] = the value that occurs most in y_neighbor 

Metrics Calculation
Calculate the accuracy by comparing y_test and y_test_predicted
Print the accuracy (Test)

Notes:

• Pandas library for csv read and shuffling
• Load csv into numpy array
• shuffle two numpy arrays together
• train test split
• distance between two numpy arrays
• numpy.argmin

Solution : LINK