So, what are we going to do?

We're essentially going to look

at basic machinery of probability theories

and actually how we can make

that into probabilistic models

where we can have good algorithms.

Of course, those will be called Bayesian networks

and we'll see how to build

them, how to do inference on them and how to use them.

Thereby, I hope to expose you to

one of the most important probability

and probabilistic reasoning techniques

we have at the moment.

So we're going to look at conditional probabilities,

and unconditional probabilities,

we're going to look at

a very important concept called independence,

which will simplify reasoning a lot.

We're going to cast this into something called Bayes' rule,

which actually is something we're going to

you implement in these Bayesian networks.

The central thing is that independence is wonderful

if you have it, except you almost never do.

And so what we're going to do is we're going to generalize this

to conditional independence,

which will be the kind of mover and shaker in this respect.

I finally managed to complete the motivation.

We can now go to the math.