Need help understanding this normalization problem!!

[jadeus]

Hi guys!

So we have a data set that includes 8 categories with a population size of 10,000. Each category has multiples values within them. A score is assigned to each category value that is equal to the population size over the number of occurrences of that value within that category (populationCount / occurrencesForThatValue).
The categories need to be normalized in such a way that applies more weight to the categories with less values in them.

The table below has the 8 categories and how many values are in each category.


The table below shows Category C as an example with its 8 values and each values' scores and normalized scores with the occurrences of each value in the population.




We have tried multiplying the total score by the normalization formula below but the numbers are not matching.



Any help is appreciated. Thanks!

 

[BigBeachBanana]

By normalization, are you referring to the normal distribution? i.e. the z-score. The formula attached at the end refers to the min-max scaling so that your data will end up in a range between 0 and 1.

The categories need to be normalized in such a way that applies more weight to the categories with less values in them.

Why more weight with less value? Shouldn't it be more value, more weight?
You defined Value=Count/Occurrence. In table 1, Categories C has a value of 7. I don't understand how table 2 relates to table 1?