ML Algorithm Implementations/Data manipulation

Question 1

What is the process of implementing Linear Regression with sklearn?

Answer 1

1. Given the dataset, Identify independant and dependant variables
2. Do EDA and Data preprocessing
3. Reshape x and y to be in right format:
   a.. x = x.reshape(-1,1)
   y = y.reshape(-1,1)
4. split your train, and test set:

x_train, x_test, y_train, y_test = train_test_split(x,y,test_size)

5. instantiate and train model

model = LinearRegression()

model.fit(x_train,y_train)

6. Get your predictions and loss values to evaluate your model

y_pred = model.predict(x_test)

mae = mean_absolute_error(y_true = y_test,y_pred = y_pred)
mse = mean_squared_error(y_true = y_test,y_pred = y_pred)
rmse = mean_squared_error(y_true = y_test,y_pred = y_pred, squared=False)

Question 2

What libraries are needed to be imported for most ML algorithms

Answer 2

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt
from sklearn.model_selection import train_test_split
from sklearn.metrics import mean_squared_error, mean_absolute_error
from sklearn.preprocessing StandardScaler

Question 3

What are the regression algorithm import lines

Answer 3

from sklearn.linear_model import LinearRegression, LogisticRegression

Question 4

Where do you import PolynomialFeatures?

Answer 4

from sklearn.preprocessing import PolynomialFeatures

Question 5

What are the import lines for KMeans?

Answer 5

from sklearn.cluster import KMeans
from sklearn.preprocessing import StandardScaler
from sklearn.metrics import confusion_matrix
from sklearn.metrics import adjusted_rand_score
import numpy as np

Question 6

What is the process of implementing Logistic Regression with sklearn

Answer 6

Identify Target and independant variables: (x,y = data.data, data.target)
Split your train and test data: (train_test_split(x,y,test_size))
scale your independant variables in both train and test data :
scaler = StandardScaler()
x_train = scaler.fit_transform(x_train)
x_test = scaler.transform(x_test)
instantiate and train your model:
model = LogisticRegression()
model.fit(x_train,y_train))
Evaluate model: model.score(x_test,y_test)

Question 7

What is the process of implementing Polynomial Regression

Answer 7

Essentially the same process as linear regression, Identify what your independant and dependant variables are, split them and reshape them.
This time, instantiate a poly_features = PolynomialFeatures(degree,include_bias)
then pass through your x through the poly_features, by doing poly_features.fit_transform(x)
now in the rest of your training and calculations, use the polynomial transformed x
y stays the same

Question 8

What is the process of implementing Kmeans

Answer 8

First get your dataset and ensure it is a classification problem
Second, Identify your dependant and independant variables and save them as such

iris = load_iris(as_frame=True)
x = iris.data
y = iris.target

third, scale your independant variables

scaler = StandardScaler()
x_scaled = scaler.fit_transform(x)

fourth, instantiate your model, then fit your model

kmeans = KMeans(n_clusters=3)
kmeans.fit(x_scaled)
labels = kmeans.labels_

fifth, visualize your data if need be, and evaluate your model using
a confusion matrix of a Rand Index

Compute the confusion matrix
conf_matrix = confusion_matrix(true_labels, labels)

Calculate purity
purity = np.sum(np.amax(conf_matrix, axis=0)) / np.sum(conf_matrix)
print(f”Cluster Purity: {purity:.2f}”)

Question 9

What is the process for reducing dimensionality using PCA in sklearn?

Answer 9

First scale the data as PCA is sensitive to input data.

Then initialize pca object, with # of desired components

pca = PCA(n_components)
x_pca = pca.fit_transform(x_scaled)

now x_scaled should have your desired component #

Question 10

What is the process for implementing a decision tree in sklearn?

Answer 10

Once you have your dependant and independant variables,

split train and test sets

Initialize and train classifier

clf = DecisionTreeClassifier()

clf.fit(x_train, y_train)

Predict:

y_pred = clf.predict(x_test)

Thats the main process

Question 11

What is the process for implementing a random forest classifier in sklearn?

Answer 11

Once you have your dependant and independant variables,

split train and test sets

Initialize and train classifier

clf = RandomForestClassifier()

clf.fit(x_train, y_train)

Predict:

y_pred = clf.predict(x_test)

Thats the main process

Question 12

What is the process for implementing a SVM classifier in sklearn?

Answer 12

Once you have your dependant and indpendant variables,

Scale your independant variable, as svm’s are sensitive to scale of data

Instantiate the model and train it

svm = SVC(kernel = “linear”)
svm.fit(x_train_scaled, y_train)

Make predictions:

y_pred = svm.predict(y_test)