This document provides an overview of machine learning, including: - Machine learning allows computers to learn from data without being explicitly programmed, through processes like analyzing data, training models on past data, and making predictions. - The main types of machine learning are supervised learning, which uses labeled training data to predict outputs, and unsupervised learning, which finds patterns in unlabeled data. - Common supervised learning tasks include classification (like spam filtering) and regression (like weather prediction). Unsupervised learning includes clustering, like customer segmentation, and association, like market basket analysis. - Supervised and unsupervised learning are used in many areas like risk assessment, image classification, fraud detection, customer analytics, and more

1. An Introduction

2. Machine Learning
User
Hey SIRI, how far is the
nearest subway?

3. User Siri
Hey SIRI, how far is the
nearest subway?
Machine Learning

4. User Siri
Decoding with the help of ML and
neural network
Apple Server
Hey SIRI, how far is the
nearest subway?
Machine Learning

5. User Siri Apple Server
Decoding with the help of ML and
neural network
Processing
Hey SIRI, how far is the
nearest subway?
Machine Learning

6. User Siri Apple Server
Decoding with the help of ML and
neural network Desired output
Hey SIRI, how far is the
nearest subway?
Machine Learning
Processing

7. User Siri Apple Server
Decoding with the help of ML and
neural network Desired output
The nearest SUBWAY
is 4km away
Hey SIRI, how far is the
nearest subway?
Machine Learning
Processing

8. What is Machine Learning?
Machine Learning is the science of making computers learn and act like humans by feeding data
and information without being explicitly programmed!

9. Machine Learning is the science of making computers learn and act like humans by feeding data
and information without being explicitly programmed!
Past Data
What is Machine Learning?

10. Analyse
Machine Learning is the science of making computers learn and act like humans by feeding data
and information without being explicitly programmed!
Data is processed
Past Data
What is Machine Learning?

11. Analyse
Train
Machine Learning is the science of making computers learn and act like humans by feeding data
and information without being explicitly programmed!
Past Data
Data is processed System Learns
What is Machine Learning?

12. Analyse
Train
Machine Learning is the science of making computers learn and act like humans by feeding data
and information without being explicitly programmed!
Past Data
System LearnsData is processed
What is Machine Learning?

13. Output
Analyse
Train
Prediction
Machine Learning is the science of making computers learn and act like humans by feeding data
and information without being explicitly programmed!
Past Data
Machine Learning makes
predictions and decisions
based on past data
System LearnsData is processed
What is Machine Learning?

14. Types of Machine Learning
Machine Learning

15. Supervised Learning
Machine Learning
Types of Machine Learning

16. Supervised Learning Unsupervised Learning
Machine Learning
Types of Machine Learning

17. Supervised Learning

18. Supervised Learning
In Supervised Learning, a model is able to
predict with the help of labeled dataset

19. Model
Supervised Learning
Known Data

20. Model
Supervised Learning
Known ResponseKnown Response
Spoon
Known Data
Knife

21. Model
New Data
Supervised Learning
Known ResponseKnown Response
Spoon
Known Data
Knife

22. Response
Spoon
New Data
Knife
Model
Supervised Learning
Known ResponseKnown Response
Spoon
Known Data
Knife

23. Types of Supervised Learning
Supervised Learning is
basically of two types
When the output variable is categorical i.e. with 2 or more classes (yes/no,
true/false), we make use of classification
Classification

24. Regression
Relationship between two or more variables where a change in one variable is
associated with a change in other variable
Supervised Learning is
basically of two types
Types of Supervised Learning
Classification When the output variable is categorical i.e. with 2 or more classes (yes/no,
true/false), we make use of classification

25. SpamFiltering
Not spam
Spam
Spam Mails
Types of Supervised Learning
Classification

26. SpamFiltering
Not spam
Spam
Learns
Spam Mails
Types of Supervised Learning
Classification

27. SpamFiltering
New mail
Not spam
Spam
Learns
Spam Mails
Types of Supervised Learning
Classification

28. SpamFiltering
New mail
Not spam
Spam
Scans the contentLearns
Spam Mails
Types of Supervised Learning
Classification

29. Classification
SpamFilteringSpam
Not spam
Spam
Categorical
Separation
New mail
Learns
Spam Mails
Not spam
Scans the content
Types of Supervised Learning

30. Weather Prediction
Types of Supervised Learning
Past Data
Regression
Temperature
Humidity

31. Weather Prediction
Types of Supervised Learning
Past Data
Regression
LearnsTemperature
Humidity

32. Regression
Weather Prediction
Types of Supervised Learning
Past Data
LearnsTemperature
Humidity
New data
Prediction

33. Weather Prediction
Types of Supervised Learning
Learns
Past Data
Temperature
Humidity
New data
Prediction
Regression

34. Areas where Supervised
Learning is used
Supervised Learning
Types of Supervised Learning

35. Areas where Supervised
Learning is used
Supervised Learning
Risk Assessment
Applications of Supervised Learning

36. Areas where Supervised
Learning is used
Supervised Learning
Image ClassificationRisk Assessment
Applications of Supervised Learning

37. Areas where Supervised
Learning is used
Supervised Learning
Fraud Detection
Image ClassificationRisk Assessment
Applications of Supervised Learning

38. Visual RecognitionFraud Detection
Areas where Supervised
Learning is used
Image ClassificationRisk Assessment
Supervised Learning
Applications of Supervised Learning

39. Unsupervised Learning

40. Unsupervised Learning
In Unsupervised Learning, the algorithm is trained
using data that is unlabeled

41. Unsupervised Learning
Known Data

42. Unsupervised Learning
Model
Pattern Recognition
Known Data

43. Unsupervised Learning
Model
Pattern Recognition
ResponseKnown Data

44. Types of Unsupervised Learning
Unsupervised Learning is
basically of two types
The method of dividing the objects into clusters which are similar between them and
are dissimilar to the objects belonging to another cluster
Clustering

45. Association Discovering the probability of the co-occurrence of items in a collection
Unsupervised Learning is
basically of two types
Types of Unsupervised Learning
Clustering
The method of dividing the objects into clusters which are similar between them and
are dissimilar to the objects belonging to another cluster

46. Suppose a telecom company wants to reduce its customer churn rate by
providing personalized call and data plans
Types of Unsupervised Learning
Clustering

47. Suppose a telecom company wants to reduce its customer churn rate by
providing personalized call and data plans
Types of Unsupervised Learning
Total Call duration
Internet Usage
Clustering

48. Suppose a telecom company wants to reduce its customer churn rate by
providing personalized call and data plans
Types of Unsupervised Learning
Total Call duration
Internet Usage
Clustering

49. A
B
C
Suppose a telecom company wants to reduce its customer churn rate by
providing personalized call and data plans
Types of Unsupervised Learning
Clustering
Total Call duration
Internet Usage
Call duration
Internetusage

50. Clustering
Call duration
Internetusage
A
B
C
Total Call duration
Internet Usage
The model segments the customers with similar traits
Suppose a telecom company wants to reduce its customer churn rate by
providing personalized call and data plans
Types of Unsupervised Learning

51. A
B
C
The members in group B have high internet usage and low call duration
and hence the company offers them the best data plans
Suppose a telecom company wants to reduce its customer churn rate by
providing personalized call and data plans
Types of Unsupervised Learning
Total Call duration
Internet Usage
Call duration
Internetusage
Clustering

52. Customer1
• Bread
• Milk
• Fruits
• Wheat
Types of Unsupervised Learning
Association

53. • Bread
• Milk
• Fruits
• Wheat
• Bread
• Milk
• Rice
• Butter
Customer1 Customer2
Types of Unsupervised Learning
Association

54. • Bread
• Milk
• Fruits
• Wheat
Customer1 Customer2 Customer3
Types of Unsupervised Learning
Association
• Bread
• Milk
• Rice
• Butter

55. • Bread
• Milk
• Fruits
• Wheat
.
If a new customer purchases
bread, he is likely to purchase
milk too
Customer1 Customer2 Customer3
Types of Unsupervised Learning
Association
• Bread
• Milk
• Rice
• Butter

56. Areas where Unsupervised
Learning is used
Unsupervised Learning
Market Basket Analysis
Applications of Unsupervised Learning

57. Areas where Unsupervised
Learning is used
Market Basket Analysis Semantic Clustering
Unsupervised Learning
Applications of Unsupervised Learning

58. Areas where Unsupervised
Learning is used
Market Basket Analysis Semantic Clustering
Delivery Store
Optimization
Unsupervised Learning
Applications of Unsupervised Learning

59. Identifying Accident
Prone Areas
Market Basket Analysis
Areas where Unsupervised
Learning is used
Semantic Clustering
Delivery Store
Optimization
Unsupervised Learning
Applications of Unsupervised Learning

60. Requires both an input and an output to
be given to the model for it to be
trained.
Supervised Learning Unsupervised Learning
• Uses known and labeled data as input • Uses unlabeled data as input
• Most commonly used unsupervised
learning algorithms are k means clustering,
hierarchical clustering, apriori algorithm
• Most commonly used supervised learning
algorithms are decision tree, logistic
regression, support vector machine
• Supervised learning has a feedback
mechanism
• Unsupervised learning has no feedback
mechanism
Recap

61. So what’s
your next step?