Python Diabetes Prediction Machine Learning Support Vector Machine Algorithm

Python tutorial diabetes prediction with machine learning Support Vector Machine Algorithm. This dataset can be downloaded from the UCI Machine Learning Repository. If you’re not familiar with the diabetes dataset, spend some time analyzing the data with a step-by-step guide on the Diabetes Dataset Analysis tutorial.

Support Vector Machine Learning Algorithm

The support vector machines model will provide an effective high dimensional spaces even when the number of dimensions is greater than the number of samples.Also, the model will use a subset of training points in the decision function (called support vectors). By doing so, it provides better memory efficiency.

Import Packages and Diabetes Data

from sklearn.svm import SVC
from sklearn.model_selection import train_test_split
import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
diabetes = pd.read_csv('diabetes.csv')

Train and Test Data

X_train, X_test, y_train, y_test = train_test_split(diabetes.loc[:, diabetes.columns != 'Outcome'],
                                                    diabetes['Outcome'], stratify=diabetes['Outcome'], random_state=66)

Build the Random Forest Classifier Model

Input:

vector = SVC()
vector.fit(X_train, y_train)
print(f"Support vector machine training set accuracy: {format(vector.score(X_train, y_train), '.4f')} ")
print(f"Support vector machine testing set accuracy: {format(vector.score(X_test, y_test), '.4f')} ")

Output:

Support vector machine training set accuracy: 1.0000 
Support vector machine testing set accuracy: 0.6510

The support vector machine gives us an accuracy score of 65.10% on the testing data but 100% on training set. This is an indicator of over-fitting and we can implement the MinMaxScaler.  The features within SVM requires the features to vary on a similar scale. We’re able to re-scale the diabetes data so that all features will be the same scale approximately.

SVM MinMaxScaler

The MinMaxScaler method transforms features by scaling each feature to a given range. Meaning this estimator scales and translates each feature individually such that it is in the given range on the training set.

Input:

from sklearn.preprocessing import MinMaxScaler
scaler = MinMaxScaler()
X_train_scaled = scaler.fit_transform(X_train)
X_test_scaled = scaler.fit_transform(X_test)
vector = SVC()
vector.fit(X_train_scaled, y_train)
print(f"Support vector machine training set accuracy: {format(vector.score(X_train_scaled, y_train), '.4f')} ")
print(f"Support vector machine testing set accuracy: {format(vector.score(X_test_scaled, y_test), '.4f')} ")

Output:

Support vector machine training set accuracy: 0.7691 
Support vector machine testing set accuracy: 0.7708