Top Page | Upper Page | Contents | About This Site | JAPANESE

Made by Normal Distribution But Not Normal Distribution

Statistics has many methods using the normal distribution and Others.

If the distribution is non-symmetric, it is difficult to judge outliers because there is the long tail.

In such case, other distributions and non-parametric are solutions. But if we can use the methods for the normal distribution, it is good.

If we find the mechanism between the real data and normal distribution, it might be a breakthrough for the data analysis.

I collect the mechanisms that I met before in my real jobs.

non-synmmetric

There are sample file to show my method to make the data in this page. But it is small size.

Made by One Set of Normal Distribution

The data n =100000, Average = 10, Standard deviation =1.

graph graph graph graph

Made by Two Sets of Normal Distribution

Not clearly, but it is the non-symmetric distribution. Right side is longer.

graph

Made by Calculated Data of Many Distributions.

1000 sets of normal distribution. N = 200, Ave. = 10 and Standard deviation = 1 for each set.

graph graph

Distribution of data created by random walk

The distribution of data created by the random walk model is far from the normal distribution, but it is still a type of distribution created from the normal distribution.
01data


<

NEXT Extreme Value Statistics