Iris Prediction Using K Nearest Neighbors

By Gautam Dhall

Iris is flower with different species. In this project we will learn how to predict the flower species using different petal size,sepal size length etc.

In this project we will learn to build our model which will help us in predicting the differnet species of a flower using different petal size,length etc. using K nearest neighbors algorithim(KNN).

So lets get started

First of all we need to install the iris dataset. For it download from the given link below

https://www.kaggle.com/uciml/iris/data#

Save the above file in CSV UTF-8 (Common Delimited)

Now , we need to import our essential libraries

Go to Command Prompt and type 

pip install numpy pandas sklearn

Numpy is used to do the matematical operations

Pandas is used to read the data set in an organized manner

sklearn in a Machine Learning library which helps in prediction.

After installing everything we are now ready to start our project. So lets start.

Step1:

Import  all the esential libraries.

import numpy as np
import pandas as pd
from sklearn.neighbors import KNeighborsClassifier

Step 2:

Read the data set using the pandas library.

data = pd.read_csv("datasets_19_420_Iris.csv")
data.head()

and you can print the dataset using data.head() in a tabular form.

We need only the first four coloumns for out training set and the last coloumn for our test set.

Step 3:

We now print the data which would require for our training data.

X = data.iloc[:,0:4]
X.head()

Similary we can read our y data set which is our test set

y = data.iloc[:,-1]
y.head()

Step 4:

Next we find shape of the dataset. The first element shows the number of data examples(number of rows) and the second element show th number of columns

Using the numpy,we find the square root of the first element of the shape.  The approx value of the square root will act as 'n' in our KNN.

data.shape
np.sqrt(data.shape[0])
knn = KNeighborsClassifier(n_neighbors = 12)

 

Step 5:

We then use the fit function. Fit measures how well a machine learning model generalizes to similar data to that on which it was trained. 

We then use the predict function.We predict our training set.

y_pred = knn.predict(X)
y.values#testing set

Step 6:

We create a Data frame where we can store our predicted data in it and print it.

y_prediction = pd.DataFrame(data = [y_pred,y.values])
y_prediction.values

Step 7:

Our model is trained and now we can predict the iris species

a = knn.predict([[1,1,1,1]])
a

 

The elements in the list are:

SepalLengthCm	SepalWidthCm	PetalLengthCm	PetalWidthCm

 Now we can give any values and predict the species.

 

The full code:

import numpy as np
import pandas as pd
from sklearn.neighbors import KNeighborsClassifier
data = pd.read_csv("iris.csv")
data.head()
X = data.iloc[:,0:4]
X.head()
y = data.iloc[:-1]
y.head()
data.shape
np.sqrt(data.shape[0])
knn = KNeighborsClassifier(n_neighbors = 12)
knn.fit(X,y)
y_pred = knn.predict(X)
y.values
y_prediction = pd.DataFrame(data = [y_pred,y.values])
a = knn.predict([[1,1,1,1]])
a

Cheers!!

Thank you