This is a very quick section: I just want to remind you about the importance of normalizing your data, making sure that your various input feature data is on the same scale, and is comparable. And, sometimes it matters, and sometimes it doesn't. But, you just have to be cognizant of when it does. Just keep that in the back of your head because sometimes it will affect the quality of your results if you don't.

So, sometimes models will be based on several different numerical attributes. If you remember multivariant models, we might have different attributes of a car that we're looking at, and they might not be directly comparable measurements. Or, for example, if we're looking at relationships between ages and incomes, ages might range from 0 to 100, but incomes in dollars might range from 0 to billions, and depending on the currency it could be an even larger range! Some models are okay with that.

If you're doing a regression, usually that's not a big deal. But, other models don't perform so well unless those values are scaled down first to a common scale. If you're not careful, you can end up with some attributes counting more than others. Maybe the income would end up counting much more than the age, if you were trying to treat those two values as comparable values in your model.

So this can introduce also a bias in the attributes, which can also be a problem. Maybe one set of your data is skewed, you know, sometimes you need to normalize things versus the actual range seen for that set of values and not just to a 0 to whatever the maximum is scale. There's no set rule as to when you should and shouldn't do this sort of normalization. All I can say is always read the documentation for whatever technique you're using.

So, for example, in scikit-learn their PCA implementation has a whiten option that will automatically normalize your data for you. You should probably use that. It also has some preprocessing modules available that will normalize and scale things for you automatically as well.

Be aware too of textual data that should actually be represented numerically, or ordinally. If you have yes or no data you might need to convert that to 1 or 0 and do that in a consistent matter. So again, just read the documentation. Most techniques do work fine with raw, un-normalized data, but before you start using a new technique for the first time, just read the documentation and understand whether or not the inputs should be scaled or normalized or whitened first. If so, scikit-learn will probably make it very easy for you to do so, you just have to remember to do it! Don't forget to rescale your results when you're done if you are scaling the input data.

If you want to be able to interpret the results you get, sometimes you need to scale them back up to their original range after you're done. If you are scaling things and maybe even biasing them towards a certain amount before you input them into a model, make sure that you unscale them and unbias them before you actually present those results to somebody. Or else they won't make any sense! And just a little reminder, a little bit of a parable if you will, always check to see if you should normalize or whiten your data before you pass it into a given model.

No exercise associated with this section; it's just something I want you to remember. I'm just trying to drive the point home. Some algorithms require whitening, or normalization, some don't. So, always read the documentation! If you do need to normalize the data going into an algorithm it will usually tell you so, and it will make it very easy to do so. Please just be aware of that!