Machine Learning

SVM Hyperparameter Tuning

You May import any type of data you want.

1. Using GridSearchCV, determine the best choice of hyperparameters out of the following possible values:

Kernel type: Linear, radial basis function

Box constraint (C): [1, 5, 10, 20]

Kernel width (gamma): 'auto','scale'

2. Report the time required to perform cross-validation via GridSearchCV. Report the mean and standard deviation of the performance metrics for the best performing model along with its associated hyperparameters. You may use the function collate_ht_results for this purpose.

Code::

#Summarizes model performance results produced during hyperparameter tuning

def collate_ht_results(ht_results,metric_keys=metric_keys,display=True):

ht_stats=dict()

for metric in metric_keys:

ht_stats[metric+"_mean"] = ht_results.cv_results_["mean_test_"+metric][ht_results.best_index_]

ht_stats[metric+"_std"] = metric_std = ht_results.cv_results_["std_test_"+metric][ht_results.best_index_]

if display:

print("test_"+metric,ht_stats[metric+"_mean"],"("+str(ht_stats[metric+"_std"])+")")

return ht_stats

UPDATE::

You can use any data set. if you need, 3 choices

#generate random data

rows, cols = 50, 5

r = np.random.RandomState(0)

y = r.randn(rows)

X = r.randn(rows, cols)

----------------------------------------

from sklearn import datasets

# FEATCHING FEATURES AND TARGET VARIABLES IN ARRAY FORMAT.

cancer = datasets.load_breast_cancer()

# Input_x_Features.

x = cancer.data

# Input_ y_Target_Variable.

y = cancer.target

# Feature Scaling for input features.

scaler = preprocessing.MinMaxScaler()

x_scaled = scaler.fit_transform(x)

---------------------------------------------

import numpy as np

X = np.array([[1, 2], [3, 4], [1, 2], [3, 4]]) y = np.array([0, 0, 1, 1])