Machine Learning

Sample Questions:
1.Obtain regression equation of Y on X and estimate Y when X=55 from the following
(i) Regression coefficients of Y on X

(ii) Regression equation of Y on X

Y–51.57 = 0.942(X–48.29 )

Y = 0.942X–45.49+51.57=0.942 #–45.49+51.57

Y = 0.942X+6.08

The regression equation of Y on X is Y= 0.942X+6.08 Estimation of Y when X= 55

Y= 0.942(55)+6.08=57.89
Find linear regression equation for the following two sets of data:

x 2 4 6 8

y 3 7 5 10

Solution:

Construct the following table:

x y x2 xy

2 3 4 6

4 7 16 28

6 5 36 30

8 10 64 80

∑x ∑y ∑x2 ∑xy

= 20 = 25 = 120 = 144

b
=
n∑xy−(∑x)(∑y)n∑x2−(∑x)2
b
=
4×144–20×254×120–400
b = 0.95
a=∑y∑x2–∑x∑xyn(∑x2)–(∑x)2

a=25×120–20×1444(120)–400
a = 1.5
Linear regression is given by:

y = a + bx

y = 1.5 + 0.95 x

Problems on confusion metric :

We have the metrics from the confusion matrix and the evaluation equations, therefore we can
easily calculate the requirements from our original question.

 True Positive (TP) a correct positive test – 30

 True Negative (TN) a correct negative test – 950
 False Positive (FP) an incorrect positive test – 15
 False Negative (FN) an incorrect negative test – 5
 Calculate the accuracy, precision, recall, sensitivity and F1 score from the matrix.

Accuracy

 number of correct predictions / total number of predictions

 30+950 / 30 + 15+ 950 + 5
 = 980/1000
 = 49/50 or 98%
Precision

 true positive / true positive + false positive

 30 / 30+15
 =30/45
 =2/3 or 66.7%
Recall (and sensitivity)

 true positive / true positive + false negative

 30 / 30+5
 =30/35
 = 0.857 or 85.7%
F1 score
 2 x (precision*recall / precision + recall)
 = 2 * (0.57/1.52)
 = 2*0.375
 =0.75 or 75%

1. .Write a note on:

a. 1 History of AI with each milestones explained

b. 2.Architectureof SOM with practical applications
2. Difference between supervised unsupervised learning
3. Give real world applications SOM , RBFNN, classification, decision tree,
linear and logistic regression.
4. Explain the following terms
SVM, margine, soft margin, hard margin, hyperplain, SVR,kernel functions
5. Explain various classification and regression metrics
6. Write a note on activation function
7. Explain the concept of overfitting in machine learning and how it can be
avoided.
8. What is a training set, validation set, and test set in machine learning?
9. What is supervised learning, and how does it differ from unsupervised
learning?
10.What are some real-world applications of regression in machine learning?
11.What is unsupervised learning, and what kind of problems does it solve?
12.Provide some practical examples of unsupervised learning applications.
13.Describe the use of AI and machine learning in autonomous systems, such
as self-driving cars.
14.What is the mathematical model of linear regression, and how is it used for
predicting continuous outcomes?
15..What is the cost function used in linear regression, and why is it important?
16.Describe the gradient descent algorithm for parameter estimation in linear
regression.
17..What is the mathematical model of logistic regression, and how is it used
for binary classification?
18.How does the logistic function (sigmoid function) transform linear
regression outputs into probabilities?
19.What is the cost function used in logistic regression, and how does it differ
from the cost function in linear regression?
20.What is multiclass classification (e.g., One-vs-All or One-vs-One
approaches)?
21.How does polynomial regression extend linear regression to fit non-linear
relationships in the data?
22.How would a credit card company use decision trees to classify transactions
as either "fraudulent" or "non-fraudulent" based on transaction amount,
merchant type, time, and location?
23.Imagine you're developing a spam filter for emails. How can a decision tree
classify emails as "spam" or "not spam" based on features such as sender,
subject line, and content keywords?
24.How can an Artificial Neural Network (ANN) be used to predict whether a
patient is likely to develop heart disease based on medical history, age,
cholesterol levels, blood pressure, and family history of heart disease?
25.What is feature scaling, and why is it important in machine learning
algorithms?
26.Explain the difference between normalization and standardization for feature
scaling.
27.Provide examples of algorithms where feature scaling is crucial for good
performance.
28.What is K-means clustering, and how is it used in unsupervised learning?
29.Describe the steps involved in the K-means clustering algorithm.
30.What are some real-world applications of K-means clustering?
31.What is overfitting, and how does it affect the performance of a machine
learning model?
32.What is underfitting, and why does it occur in machine learning models?
33.How can regularization techniques (e.g., L1 and L2 regularization) help
prevent overfitting?
34.What is the difference between Lasso (L1) and Ridge (L2) regularization?
35.What are the similarities and differences between biological neurons and
artificial neurons?
36.What are the key historical milestones in the development of ANNs?
37.Explain the concept of an activation function in an ANN and its role in
introducing non-linearity.What are some common activation functions used
in ANNs?
38.What is a perceptron, and how does it function as a basic neural network
model?
39.What is Multilayer percptron model
40.Explain the process of forward propagation in a multi-layer perceptron.
41.What is a self-organizing map (SOM), and how does it function in
unsupervised learning and How does the architecture of a SOM differ from
that of other neural networks?
42.Explain the mathematical formulas used for normalization (min-max
scaling) and standardization (z-score scaling).In what range do the
transformed values lie for normalization and standardization?
43.What is a confusion matrix, and how is it used in classification problems?
Explain the components of a confusion matrix: True Positive (TP), True
Negative (TN), False Positive (FP), and False Negative (FN)
44.Name and describe the main features of Genetic Algorithms (GA).
45.What is a genetic algorithm (GA) and how is it inspired by biological
evolution?
46.Describe the process of 'selection' in genetic algorithms.
47.Explain 'crossover' and 'mutation' operations in genetic algorithms.
48.What is a 'fitness function' in the context of a genetic algorithm?
49.What are the typical stopping conditions for a GA?
50.What is Roulette wheel Selection?
51.Compare single point and two point cross over?
52.What is Uniform Crossover ?
53.What is the difference between mutation and crossover?