Kickstart Your Journey with SAP Artificial Intelligence: Part 2 – What is Machine Learning

Introduction

In our previous blog, we talked about how Artificial Intelligence (AI) helps a robot find your lost toy by thinking like a human. Now, we’re going to take it a step further and explore Machine Learning (ML), which is like teaching the robot how to learn on its own. Let’s dive into it in a fun and simple way!

This 7 part blog series is designed for anyone who wants to start learning about SAP AI. No matter what your job is or what you do, understanding the basics of Generative AI how it connects to SAP is a smart choice.

In this blog series, we’ll break down SAP AI into easy-to-understand parts, covering one step at a time.

Part 1 — Introduction to AI 

Part 2 — What is Machine Learning? [Current Blog]

Part 3 — Basics of Deep Learning

Part 4 — Getting Started with Generative AI

Part 5 — What Are Large Language Models (LLMs)?

Part 6 — Prompt Engineering: How to Communicate with AI

Part 7 — Introduction to SAP AI

Machine Learning (ML) — From a Kid’s Perspective

In our previous blog, we discussed how AI can help a robot find your lost toy. Now, let’s explore how we can teach the robot to recognize different types of toys.

Imagine we want the robot to be able to identify various toys — like cars, dolls, and building blocks. Here’s how we can do it:

1. Show and Tell: We start by showing the robot lots of pictures of different types of toys. For example, we show it many pictures of cars, dolls, and building blocks, each labeled with the toy’s name.

2. Teach the Robot: We need to train the robot to recognize these toys based on their size, color, shape, and other features. This is like teaching a friend what your favorite toys look like so they can help you find them.

3. Recognize New Toys: Once the robot has seen many pictures of each type of toy, it can start to recognize them in real life, even if they don’t look exactly like the pictures. For example, even if you have a new toy car that looks a bit different from the ones we showed, the robot should still be able to identify it as a toy car.

4. Practice Makes Perfect: The more pictures of toys the robot sees, the better it becomes at recognizing them. This is because it learns to pick out common features and differences between toys, helping it make better guesses.

This process of teaching the robot with lots of examples is known as Machine Learning.

In summary, Machine Learning is:

• A subset of Artificial Intelligence: It’s one way of making machines smart.
• Learning from Examples: It involves giving machines lots of data (like pictures of toys) so they can learn and make decisions based on that data.

So, just like you learn to recognize different toys by seeing and playing with them, robots learn to identify toys by looking at many examples. The more they see, the smarter they get!

Types of Machine Learning

Now, let’s dive into the different ways robots and computers can learn from data. Machine Learning can be categorized into three main types:

1. Supervised Learning
2. Unsupervised Learning
3. Reinforcement Learning

Each type of Machine Learning serves a unique purpose and uses different methods to learn from data. Let’s explore each type in detail.

Supervised Learning

Remember when we taught the robot to identify various toys by showing it lots of pictures of different types of toys, each labeled with the toy’s name? This method is known as Supervised Learning.

In Supervised Learning:

• We Provide the Answers: Just like you told the robot which toy is a car, doll, or block, we provide the machine with examples that are labeled with the correct answers.
• The Machine Learns: The machine uses these labeled examples to learn how to identify or predict new, unlabeled data.

Here’s how Supervised Learning works in the real world:

• Email Spam Filtering: Classifying emails as spam or not spam based on features from the email content and sender.
• Image Classification: Identifying objects in images, such as recognizing animals or detecting objects in self-driving cars.
• Facial Recognition: Verifying individuals based on facial features, used in security systems or unlocking devices.
• Financial Fraud Detection: Spotting potentially fraudulent transactions by analyzing patterns in financial data.
• Speech Recognition: Converting spoken language into text, like in voice assistants such as Siri or Google Assistant.

Unsupervised Learning

Let’s think about how kids learn to group their classmates on the first day of school. They notice differences and similarities among their classmates without anyone explicitly telling them how to categorize them. This is similar to Unsupervised Learning.

In Unsupervised Learning:

• No Labels Provided: The machine is given data without any labels or categories. It must find patterns and group similar items on its own.
• Pattern Discovery: The machine identifies similarities and differences in the data, grouping similar items together based on the patterns it discovers.

For example:

• Clustering Customer Segmentation: Businesses use unsupervised learning to group customers based on purchasing behavior, helping tailor marketing efforts.
• Anomaly Detection in Cybersecurity: Identifying unusual patterns in network traffic to flag potential security threats.
• Recommendation Systems: Suggesting products or content based on patterns in user behavior, like how Netflix recommends movies.

Reinforcement Learning

Imagine teaching a dog a new trick. You give the dog a treat when it performs the trick correctly and no treat when it doesn’t. Over time, the dog learns to do the trick more often to get more treats. This is similar to Reinforcement Learning.

In Reinforcement Learning:

• Learning by Trial and Error: An agent (like a robot or computer program) learns to make decisions by receiving rewards for good actions and penalties for bad ones.
• Interactive Learning: The agent interacts with its environment, learns from the outcomes of its actions, and adjusts its behavior to maximize rewards.

For example:

• Game Playing: AlphaGo, developed by DeepMind, uses reinforcement learning to play the board game Go at a superhuman level, learning the best strategies through practice.
• Self-Driving Cars: Reinforcement learning helps these cars learn how to navigate traffic and make driving decisions by simulating various driving conditions.
• Algorithmic Trading: Making trading decisions based on historical and real-time market data to develop effective trading strategies.

Summary

Machine Learning is a branch of Artificial Intelligence that allows computers to learn from data and experiences, without being explicitly programmed for every task. Whether it’s recognizing toys, understanding voice commands, or helping doctors analyze medical images, Machine Learning is becoming an integral part of our everyday lives.