Welcome everybody.

So today I want to answer a couple of questions that you have been asking over the last couple

of weeks and they are of course all regarding our lecture pattern recognition.

Some of you had a bit of trouble following the short videos because we have like one

proof or one chain of argument in a video but sometimes it's difficult to preserve

the entire overview over the complete lecture.

For these reasons I prepared a short video here where we are summarizing the entire content

of the lecture and we give a quick summary of everything that will happen in this course.

Okay, so pattern recognition the big picture.

Of course you want to learn Python and machine learning and pattern recognition, all of those

things.

So I guess this is also one of the reasons why you are watching this video.

But then again some of you might not be so eager to learn the math.

So I think this is still a set of skills that you may want to be able to acquire because

many of you do this course here because they want to work as a data scientist or a researcher

as an artificial intelligence later on.

So I think the math is still something that you should be aware of and believe me this

math is not too hard.

Of course machine learning is kind of difficult to access if you don't know the math.

This is why we also provide now extra videos where we explain basics of math that are fundamental

to follow our course.

But let's say it's not just that you have to be good at math but it's actually all about

function fitting and matrix multiplication.

So these are essentially the key techniques that you need for this course and also for

a career in machine learning.

So you don't have to follow all of the topics of math.

There's a couple of things that you have to know by heart and these are also the ones

that we will emphasize in the additional videos that we now provide.

Now what's the big picture?

So our class is of course pattern recognition.

So here is our pattern recognition cloud and these are the topics that we are discussing.

So we start with an introduction.

We start with classification and introduce why we need this.

So this is essentially in order to be able to derive a decision from some input data

and we do that with the pattern recognition pipeline.

So the pipeline starts with the sensor.

The sensor is digitized.

Then we extract features.

And finally after we have found suitable features we actually go ahead and then train a classifier.

Now this class is mainly about the classifiers and not so much about the feature extraction.

We have the other course introduction to pattern recognition where we talk more about the sensors

and suitable features for certain tasks.

Then we also introduce evaluation measures but the evaluation measures we only talk about

very little so don't be afraid this is not too relevant if you want to prepare for example

for our examination.

But of course you need to know about classifiers and how to evaluate them because this will

be key for your career.

So we have only a very coarse overview over different evaluation measures here.

Then we go ahead to the basics.

We review the base theorem and have a small example about that.