Okay, so welcome everybody to our lecture introduction to software engineering.

And today we already have the last of our lectures, where we will be looking into the

implications of machine learning into the field of software engineering.

And then we will also have a short discussion about the exam.

So I will record, of course, the part with the introduction to machine learning and software

engineering.

And then when it comes to the questions about the exam, I will stop the recording such that

everybody can ask questions without having to be afraid about the recording.

All right.

So yeah, let's go into our topic.

What is machine learning?

And how can this help us to make better software?

And this is really a very recent development.

So I'm actually quite excited that as the pattern recognition lab are able to include

that already in our teaching materials, because you will see that machine learning in particular

foundation models and large language models will reshape the world of software engineering

quite considerably.

And therefore, I think you should have heard about this early on and already know about

things that can, yeah, what they are good for, what they can be used for, and probably

also some hints about what they are not very good for.

So let's dive into the topic.

So machine learning, well, it's been used across many, many industries.

So you can see there's chatbots, autonomous driving is around.

Tesla and Elon Musk keep announcing it as a final product within the next six months.

And I think this has been going on over the past five years.

When you ask about when will autonomous driving be ready, it's the people will answer in six

months. But there's also, of course, many other applications.

The chat GPT has been quite revolutionizing and also programming with large language models

is now something that is possible, but it's also applied for user recommendation.

The Netflix challenge was something that has been tackled with machine learning in a couple

of years ago. Netflix wasn't able to recommend very well what if you like the show, what

other shows you might want to watch.

But this has been solved.

Apple is heavily investing in image processing, facial recognition.

Your smartphone does a lot of this stuff, but also the tools within your Apple computer

are quite heavily influenced by machine learning methods.

For example, the Apple mankind can also automatically categorize your emails.

It's not just spam versus no spam, but they also have topic detection and help you to

put them into the right folders.

And of course, with our relations here in Alan, we do a lot of medical applications

where we employ machine learning methods.

So this is really something that has had a tremendous boost in the past few years.

So I would say over the scope of the last 10 years, it has tremendously been increasing.

I remember that pattern recognition was a lecture where we had something like masters

specialized at most 30 students.

Now we have a pattern recognition more than a thousand students every time we offer the

course. So it's really amazing how much attraction is also caused by this topic.

So, yeah, what is machine learning?

Well, you write an algorithm for a task and the task must be well understood and then

solvable using equations.