ARTIFICIAL INTELLIGENCE

1. INTRODUCTION TO AI ON AZURE

1. Introduction to AI
What is AI?
• AI is the creation of software that imitates human-behaviors and human-capabilities.

1.1 Five key workloads of AI

i) Machine-Learning
• We "teach" a computer-model to make prediction and draw conclusions from data.
ii) Anomaly-Detection
• The capability of software to automatically detect errors or unusual activity in a system.
iii) Computer-Vision
• The capability of software to interpret the world visually through cameras, video and images.
vi) Natural Language Processing
• The capability of software to interpret written or spoken language, and respond in kind.
v) Knowledge Mining
• The capability of software to extract information from large volumes of unstructured-data.
• The extracted-information is used to create a searchable-knowledge-store.

Course – 1 Azure AI Fundamentals (AI-900)

1-1
ARTIFICIAL INTELLIGENCE
2. Understand Machine-Learning (ML)
2.1 How Machine-Learning works?
• The machines learn from data.
• In today's world, huge volumes of data is created.
For e.g. Text messages, emails, social media posts, photographs and videos.
• Data is also created by millions of sensors in homes, cars, cities and factories.
• Data-scientists can use the data to train ML-models.
• The models can make predictions and inferences based on the relationships in the data.
• For example,
How to identify different species of flower using a phone app.
• The below procedure is followed:
1) Scientist collects data on flower-samples.
2) Scientist labels the samples with the correct species.
3) The labeled-species is processed using an algorithm.
The algorithm finds relationships between
i) Features of the samples and
ii) Labeled-species.
4) The results of the algorithm are captured in a model.
5) When new sample is found, the model can identify the correct species label.

2.2 Machine-Learning in Microsoft-Azure

• Microsoft-Azure provides the Machine-Learning (ML) service.
• ML-service is a cloud-based platform for creating, managing, and publishing ML-models.

Course – 1 Azure AI Fundamentals (AI-900)

1-2
ARTIFICIAL INTELLIGENCE
3. Understand Anomaly-Detection (AD)
• Anomaly-detection analyzes data over time and identifies unusual changes.
• For example,
How anomaly-detection can be used in the racing car scenario?
• The below procedure is followed:
1) In the car, sensors measure telemetry such as engine-revolutions, brake-temperature etc.
2) The model is trained to understand expected-range in the telemetry-measurements over time
3) If a measurement occurs outside of the normal expected-range, the model reports an anomaly.
This anomaly can be used to alert the race-engineer to fix the issue.

Anomaly-Detection in Microsoft-Azure
• AD service provides an API that developers can use to create anomaly-detection solutions.

Course – 1 Azure AI Fundamentals (AI-900)

1-3
ARTIFICIAL INTELLIGENCE
4. Understand Computer-Vision (CV)
• Computer-Vision deals with visual-processing.
• For example,
The Seeing app is designed for the blind and low vision community.
The Seeing app uses the power of AI to describe nearby people, text and objects.
4.1 Computer-Vision Models And Capabilities
• Most Computer-Vision solutions are based on ML-models.
• The model is trained with visual-input from cameras, videos, or images.
• The common tasks are:
1) Image Classification
• The models are trained to classify images based on their contents.
2) Object Detection
• The models are trained to
 classify individual objects within an image and
 identify location of objects with a bounding-box.
3) Semantic Segmentation
• In the image, individual pixels are classified based on the object to which they belong.
4) Face Detection, Analysis, And Recognition
• Face detection is used to locate human faces in an image.
5) Optical Character Recognition (OCR)
• OCR is used to detect and read text in images.
• You can use OCR to
 read text in photographs or
 extract information from scanned documents such as letters, invoices, or forms.

4.2 Computer-Vision Services In Microsoft-Azure

Course – 1 Azure AI Fundamentals (AI-900)

1-4
ARTIFICIAL INTELLIGENCE
5. Understand Natural Language Processing (NLP)
• NLP deals with creating software that understands written and spoken language.
• Four Tasks of NLP software:
i) Analyze and interpret text in documents, email messages and other sources.
ii) Interpret spoken language and create speech responses.
iii) Automatically translate spoken or written phrases between languages.
iv) Interpret commands and determine appropriate actions.
• For example,
Starship Commander is a VR game.
The game uses NLP to enable players to interact with in-game characters and starship systems.

5.1 Natural Language Processing In Microsoft-Azure

6. Understand Knowledge Mining (KM)

• KM involves extracting information from large volumes of unstructured-data.
• The extracted-information is used to create a searchable-knowledge-store.

6.1 Knowledge Mining In Microsoft-Azure

• Azure Cognitive Search (ACS) is one of the knowledge mining solution.
• ACS is a private, enterprise, search solution that has tools for building indexes.
• Then, the indexes can be used for internal only use.
• ACS can use the built-in AI capabilities such as
i) Image processing
ii) Content extraction and
iii) Natural language processing.

Course – 1 Azure AI Fundamentals (AI-900)

1-5
ARTIFICIAL INTELLIGENCE
7. Challenges And Risks With AI

8. Understand Responsible AI
• AI software development is guided by 6 principles.
i) Fairness
• The systems should treat all people fairly.
• For example,
Suppose you want to create a model for loan approval application.
The model should predict whether the loan should be approved or denied without bias.
ii) Reliability And Safety
• The systems should perform reliably and safely.
• For example,
` AI-system for an autonomous-vehicle.
• Unreliability in this kind of system can result in substantial risk to human life.
• The system must be subjected to rigorous testing before release.
iii) Privacy And Security
• The systems should be secure and respect privacy.
• The models rely on data, which may contain personal details that must be kept private.
iv) Inclusiveness
• The systems should enable everyone and engage people.
• The system should bring benefits to all parts of society, regardless of physical ability or gender
v) Transparency
• The systems should be understandable.
• Users should be made fully aware of
1) What is the purpose of the system?
2) How the system works?
3) What limitations may be expected?
vi) Accountability
• People should be made accountable for using the system.
• Developers of the system should work within a framework of organizational principles.

Course – 1 Azure AI Fundamentals (AI-900)

1-6