Outlining the business case solution

We know what we are supposed to do.

Let’s see how we will do it since we are working with real life data.

We must pre process it pre processing is important for machine learning in a data science team.

There may be a person whose sole job is to prepare data sets for analysis.

This is why we would like to show you a couple of common techniques here to create a machine learning

algorithm from scratch.

We need three important steps.

The first one is to balance the data set.

We will explain that in a separate lecture as it needs extra attention.

The second step is to divide the data set into three parts training validation and test.

We already know why we must do it.

Now we will see how to do it from scratch.

We will save the newly created sets in a tensor friendly format.

The good old NPC.

Finally we will create the machine learning algorithm.

The code is 90 percent the same as the one in the amnesty example.

Here is the true power of tensor flow.

We will use the same structure to create a different model which will be equally powerful.

By the end you can take any data set and repeat this operation using the code to create deep neural

networks for countless problems.

Ready.

Okay see you in the next video.