Machine learning was discussed including definitions, types, and examples. The three main types are supervised, unsupervised, and reinforcement learning. Supervised learning uses labeled training data to predict target variables for new data. Unsupervised learning identifies patterns in unlabeled data through clustering and association analysis. Reinforcement learning involves an agent learning through rewards and penalties as it interacts with an environment. Examples of machine learning applications were also provided.

1. LECTURE 6
Machine Learning
Instructor : Yousef Aburawi
Cs411 -Artificial Intelligence
Misurata University
Faculty of Information Technology
Spring 2022/2023

2. AI&ML&DL
1-2

3. What is Machine Learning
 “Learning is any process by which a system improves
performance from experience.”
- Herbert Simon
 “Machine Learning is the study of algorithms that improve their
performance P at some task T with experience E.”
- Tom Mitchell

4. Types of Machine Learning
 Supervised (inductive) learning
 Given: training data + specified target variable or attribute
 Predict: the values of the target variable on unseen/test data
 Examples: classification (predicting a categorical target variable);
regression (predicting a numeric/continuous target variable)
 Unsupervised learning
 Given: training data (without any target variable)
 Identify previously unknown patterns in the data
 Examples: clustering; association analysis
 Reinforcement learning
 Given: Rewards/feedback from sequence of actions
 Incrementally learn/optimize a model to maximize utility

5. examples of tasks best solved by ML
 Recognizing patterns / Classifying objects
 Facial identities or facial expressions
 Handwritten or spoken words
 Medical / satellite images analysis
 Text/document categorization
 Generating patterns
 Generating images or motion sequences
 Generating spoken / written narratives
 Recognizing anomalies
 Unusual credit card transactions
 Unusual patterns of sensor readings in a nuclear power plant
 Prediction
 Future stock prices or currency exchange rates
 Predicting customer preferences on items

6. Many application areas
 Web search
 Text / Web mining
 Recommender systems / Personalization
 Computational biology
 Computational Finance
 E-commerce / Target marketing
 Market segmentation and focalization
 Space exploration
 Robotics
 Autonomous vehicles
 Information extraction
 Social network analysis
 Natural language understanding
 Fraud detection
 ….

7. Machine Learning Terms that are commonly used:
 Labelled data: It consists of a set of data, an example would
include all the labelled cats or dogs images in a folder or all the
prices of the house based on size, etc.
 Classification: Separating into groups having definite values
Eg. 0 or 1, cat or dog or orange etc.
 Regression: Estimating the most probable values or
relationship among variables. Eg. estimation of the price of the
house based on size.
 Association: Discovering interesting relations between
variables in large databases where the connections that found
are crucial.
1-7

8. Machine Learning Terms that are commonly used:
 Clustering: Grouping related examples, particularly during
unsupervised learning. Once all the examples are grouped, a
human can optionally supply meaning to each cluster.
 Agent: In reinforcement learning, the entity that uses a policy
to maximize expected return gained from transitioning between
states of the environment.
 Model: The representation of what a machine learning system
has learned from the training data.
 Baseline: A model used as a reference point for comparing
how well another model (typically, a more complex one) is
performing.
1-8

9. Machine Learning Terms that are commonly used:
 Generalization: Refers to your model’s ability to make correct
predictions on new, previously unseen data as opposed to the
data used to train the model.
 Inference: Often refers to the process of making predictions by
applying the trained model to unlabeled examples.
 Convergence: Often refers to a state reached during training
in which training loss and validation loss change very little or
not at all with each iteration after a certain number of iterations.
1-9

10. Machine Learning Terms that are commonly used:
 Overfitting: Creating a model that matches the training data so
closely that the model fails to make correct predictions on new
data.
 Fine tuning: Perform a secondary optimization to adjust the
parameters of an already trained model to fit a new problem.
Fine tuning often refers to refitting the weights of a trained
unsupervised model to a supervised model.
 Feedback loop: In machine learning, a situation in which a
model’s predictions influence the training data for the same
model or another model.
1-10

11. Supervised Learning
1-11

12. Supervised Learning
 Supervised learning is task driven. You need to guide the
machine to perform a certain task. Dataset act as a teacher here
and its role is to train the machine. Once the machine gets
properly trained, it can start making predictions or decisions
whenever new data comes. Supervised Machine Learning
searches for patterns within the value labels which are assigned
to a dataset.
 For example, multiple images of a cat, dog, orange, apple, etc.
are labelled and fed into the machine for training and the machine
must identify the same. Just like a human child is shown a cat
and told so, when it sees a completely different cat among others
1-12

13. Supervised Learning Key Points:
• Regression and classification problems are mainly solved in
supervised learning.
• Labelled data is used for training here.
• Popular Algorithms: Linear Regression, Logistic
Regression, Support Vector Machines (SVM), Neural
Networks, Decision Tree, Random Forest, KNN, Naive Bayes,
Nearest Neighbor, etc.
• It is mainly used in Predicting Modelling.
1-13

14. Supervised Learning: Regression

15. Supervised Learning: Classification

16. Supervised Learning: Classification

17. Classification Example: Spam Filter
 Input: an email
 Output: spam/ham
 Setup:
 Get a large collection of example emails, each
labeled “spam” or “ham”
 Note: someone has to hand label all this data!
 Want to learn to predict labels of new, future
emails
 Features: The attributes used to make the
ham / spam decision
 Words: FREE!
 Text Patterns: $dd, CAPS
 Non-text: SenderInContacts
 …
Dear Sir.
First, I must solicit your confidence in
this transaction, this is by virture of its
nature as being utterly confidencial and
top secret. …
TO BE REMOVED FROM FUTURE
MAILINGS, SIMPLY REPLY TO THIS
MESSAGE AND PUT "REMOVE" IN THE
SUBJECT.
99 MILLION EMAIL ADDRESSES
FOR ONLY $99
Ok, Iknow this is blatantly OT but I'm
beginning to go insane. Had an old Dell
Dimension XPS sitting in the corner and
decided to put it to use, I know it was
working pre being stuck in the corner,
but when I plugged it in, hit the power
nothing happened.

18. Classification Example: Digit Recognition
 Input: images / pixel grids
 Output: a digit 0-9
 Setup:
 Get a large collection of example images, each labeled with a
digit
 Note: someone has to hand label all this data!
 Want to learn to predict labels of new, future digit images
 Features: The attributes used to make the digit decision
 Pixels: (6,8)=ON
 Shape Patterns: NumComponents, AspectRatio, NumLoops
 …
0
1
2
1
??

19. Designing a Supervised Learning System
 Choose the training data
 Choose exactly what is to be learned
 i.e. the target function
 Choose how to represent the target function
 Choose a learning algorithm to infer the target function
from the experience

20. Supervised Learning Summary: 3 Step Process
 1. Model construction (Learning):
 Each record (instance, example) is assumed to belong to a predefined class, as determined by one of the
attributes
 This attribute is call the target attribute
 The values of the target attribute are the class labels (classification) or numeric values (regression)
 The set of all instances used for learning the model is called training set
 The model may be represented in many forms: decision trees, probabilities, neural networks, ….
 Tune hyperparameters (typically through cross-validation)
 2. Model Evaluation (Accuracy):
 Estimate accuracy rate or error rate of the model based on a test set
 The known labels of test instances are compared with the predicted labels/values from model
 Test set is independent of training set otherwise over-fitting will occur
 3. Model Use (Classification):
 The model is used to classify unseen instances (e.g., to predict the class labels for new unclassified
instances)

21. Un-Supervised Learning
1-21

22. Un-Supervised Learning
 Unsupervised learning is data driven. This model learns
through observations and identifies structures in the data. Once
the model is equipped with the dataset, it automatically
identifies patterns and relationships in the dataset by creating
clusters in it and it requires to describe its structure and make
complex data look simple and highly organized for analysis.
 For example, the images or the inputs given are grouped
together in unsupervised learning and insights on the inputs
can be found here (which is the most of the real world data
available).
1-22

23. Un-Supervised Learning Key Points:
• Unsupervised learning is used for Clustering
problems(grouping), Anomaly Detection (in banks for
unusual transactions) where there is a need for finding
relationships among the data given.
• Unlabeled data is used for training here..
• Popular Algorithms: Apriori algorithm, k-means
clustering, Association rule.
• It is mainly used in Descriptive Modelling.
1-23

24. Un-Supervised Learning

25. Un-Supervised Learning

26. Un-Supervised Learning

27. Reinforcement Learning
1-27

28. Reinforcement Learning
 Reinforcement Learning learns to react to an environment.
They interact with the environment and find out what could be
the best outcome. The machine learns from past experience
and tries to capture the best possible knowledge to make
accurate decisions based on the feedback received. This type
of machine learning typically follows a hit & trial method. With
time, the algorithm changes its strategy to learn automatically.
The agent will either get rewarded or penalized for their
prediction. The model trains itself with the positive rewards it
gains.
1-28

29. Reinforcement Learning Key Points:
• Basic reinforcement is modeled as Markov Decision Process
• The most popular algorithms used here are Q-Learning, Deep
Adversarial Networks.
• Computer playing board games such as Chess and GO, Self-
driving cars are using this type of machine learning.
1-29

30. Basic Comparison
1-30

31. Readings
 Chapters 19 of Textbox.
1-31

32. The End