AI4FS Workshop

Introduction to Artificial Intelligence

Engr. Dr. Jibril Abdullahi Bala
Senior Research Fellow,
Summit University, Offa
Faculty Member,
Department of Mechatronics Engineering,
School of Electrical Engineering and Technology,
Federal University of Technology Minna.
Artificial Intelligence

 The term "Artificial Intelligence" was coined by John McCarthy

 AI is a sub field of computer science that aims at building computer systems that can
perform tasks that normally require human intelligence.
 The main goal of AI has been developing computer systems that equal or exceed
human intelligence.
 The field of AI research was born at a workshop at Dartmouth College in 1956.
 AI-based machines are intended to perceive their environment and take actions that
optimize their level of success.

8/23/2024 2
Intelligent Systems

 Intelligence is the ability of a system to learn and understand, to solve problems, and
to make decisions.
 Intelligent systems are systems which can exhibit human intelligence by applying
artificial intelligence approaches.

8/23/2024 3
Types of AI

 Strong AI: This involves building systems that can act and think intelligently like
people do. This approach simulates human reasoning and cognition in general tasks
 Examples of strong AI are DeepMind and the Human Brain Project.

 Weak AI: Here, the focus is on AI systems that perform specific tasks accurately and
correctly.
 Examples of Weak AI are Chatbots.

 Currently, AI is between strong and Weak e.g NLP (NLG and NLU)

8/23/2024 4
Multidisciplinary Applications of AI

 AI research uses techniques from many fields, such as linguistics, economics, and psychology.
 These techniques are used in applications, such as control systems, natural language processing,
facial recognition, speech recognition, business analytics, pattern matching, and data mining.
 Transportation: Autonomous Vehicles, UAVs, UGVs
 Home Automation and services: Vacuum robots, delivery drones, personal assistants
 Healthcare: Prediction and diagnosis, surgical robots
 Education: Interactive machines
 Public Safety and Security: Surveillance Drones, algorithms for fraud detection
 Manufacturing Industry: Automated processes
 Agriculture: Precision farming
 Banking: Forecasting, Stock prediction
 Oil and Gas: Pipeline Monitoring, spillage detection

8/23/2024 5
Multidisciplinary Applications of AI (Cont.)

8/23/2024 6
 Machine Learning

 The term “machine learning” was first introduced by Arthur Samuel in 1959.
 He defined it as the “field of study that gives computers the ability to learn without being explicitly
programmed”.
 If a computer program can improve how it performs certain tasks based on past experiences, then
it has learned. This differs from performing the task always the same way because it has been
programmed to do so.
 The learning process improves the so-called “model” over time by using different data points
(training data). The evolved model is used to make future predictions.

23/08/2024 Artificial Intelligence / Machine Learning and Applications 7

 Categories of Machine Learning

 Supervised Learning
 Dataset contains inputs and target outputs. Eg. ANN, SVM, DT

 Unsupervised Learning
 Data contains only inputs and no outputs. Objective is to find structure in the data. Eg. K-means Clustering

 Semi-Supervised Learning
 Falls between supervised and unsupervised. Some datasets have labels and some do not

 Reinforcement Learning
 Software agents take actions only to maximise some sort of reward

23/08/2024 Artificial Intelligence / Machine Learning and Applications 8

 What is Supervised Learning ?

 This is a type of machine learning that involves using labelled data to train the
model and predict new labels for unseen input data.
 Here, the dataset contains inputs and target outputs
 In supervised machine learning, input data (also known as training examples)
comes with a label, and the goal of learning is to predict the label for new,
unforeseen examples.
 In practice, the problems that are solved by using supervised learning are
grouped into either regression or classification problems.
 Classification is the task of predicting a discrete class label, such as “black,
white, or gray” .
 Regression is the task of predicting a continuous quantity, such as “weight”,
“probability” and “cost”.

23/08/2024 Artificial Intelligence / Machine Learning and Applications 9

 Unsupervised Learning
 This type of machine learning involves detecting patterns and relationships between
data without using labelled data.
 One of the most popular types of unsupervised learning is clustering algorithms.
 Clustering algorithms are algorithms that discover how to split the data set into a
number of groups such that the data points in the same groups are more similar to
each other compared to data points in other groups.

23/08/2024 Artificial Intelligence / Machine Learning and Applications 10

Semi-Supervised Learning

 This type of machine learning technique falls between supervised and unsupervised
learning.
 It includes some labelled data with a large amount of unlabelled data.
 Many real practical problems fall into this category of machine learning where you
have little labelled data and the rest of the data is unlabelled.
 In semi-supervised learning, you try to get the best out of your unlabelled data.
 A popular technique of Semi-Supervised Learning is Pseudo-Labelling.

23/08/2024 Artificial Intelligence / Machine Learning and Applications 11

Pseudo-Labelling

 This technique aims at providing approximate labels to unlabelled data.

 The steps for Pseudo-Labelling are as follows:

1. Use labelled data to train a model

2. Use the model to predict labels for the unlabelled data
3. Use the labelled data and the newly generated labelled data to create a model

23/08/2024 Artificial Intelligence / Machine Learning and Applications 12

Reinforcement Learning

 Reinforcement Learning uses trial and error (a rewarding approach)

 The algorithm finds a relationship between the goal and the sequence of events that
leads to a successful outcome.

 Examples include:
 Autonomous Vehicles
 Robot path finding

23/08/2024 Artificial Intelligence / Machine Learning and Applications 13

Terminologies

 Model: Mathematical function that represents a relationship or mapping between a set

of inputs and a set of outputs.
 Eg: 𝑓 𝑥 = 𝑥 2

 Machine Learning Algorithm: technique through which the system extracts useful
patterns from historical data. These patterns can be applied to new data.
 Algorithm: An unambiguous specification of how to solve a class of problems.
Algorithms can perform calculation, data processing, and automated reasoning tasks.

23/08/2024 Artificial Intelligence / Machine Learning and Applications 14

Terminologies (cont.)

 Dataset: A collection of data. Most commonly a data set corresponds to the contents
of a single database table, or a single statistical data matrix, where every column of
the table represents a particular variable, and each row corresponds to a given
member of the data set in question. The data set lists values for each of the
variables, such as height and weight of an object, for each member of the data set.
 Training: A training dataset is a dataset of examples used during the learning process
and is used to fit the parameters
 Testing: A test set is therefore a set of examples used only to assess the
performance
 Validation: A validation dataset is a dataset of examples used to tune the architecture
of a classifier.

23/08/2024 Artificial Intelligence / Machine Learning and Applications 15

Guide to AI Applications

AI

Expert Machine Deep

Optimisation
Systems Learning Learning

CNN,
GA, PSO, ANN, DT,
Fuzzy Logic RCNN,
ACO SVM
YOLO

8/23/2024 16