What is Bootstrapping (In Statistics)?
What is Bootstrapping (In Statistics)?
Bootstrapping is an interesting process or technique of assigning measure of accuracy. Depended upon calculation, Bootstrapping can be used to any statistic to measure estimation.
Definition
According to the Cambridge dictionary of statistics – “A confidence interval is a range of values, calculated from the sample observations that are believed, with a particular probability, to contain the true parameter value. A 95% confidence interval, for example, implies that were the estimation process repeated again and again, then 95% of the calculated intervals would be expected to contain the true parameter value.”
Reason for the name
The term has been derived from the expression "lifting oneself up by one's own bootstraps". This refers to raising oneself up by one's own means. This is also called bootstrap funding.
Example
Let’s understand this with an interested example of a candy factory. Knowing that the candy bars have mean weight. Since it is not feasible to weigh each candy that is produced in the factory, we use sampling techniques and randomly choose 100 candies. We calculate the mean of these 100 candies, and say that the population mean falls under a margin of error.
Suppose that after a couple of months, we want to know the mean candy weight on the day that we sampled the production line (with greater accuracy - or less of a margin of error).
We simply cannot use candies being produced today, as too many elements are now being used like; sugar, different flavours of milk, cocoa beans, different atmospheric conditions, different employees etc. All that we have from the day that we are curious about are the 100 weights.
When to use bootstrapping?
We can simply explain bootstrapping taking simple random sample from the observed sample of same size but with a replacement. This would mean each observation has equal probability of selecting in the sample.
We can use bootstrapping for inference and
Bootstrapping is an interesting process or technique of assigning measure of accuracy. Depended upon calculation, Bootstrapping can be used to any statistic to measure estimation.
Definition
According to the Cambridge dictionary of statistics – “A confidence interval is a range of values, calculated from the sample observations that are believed, with a particular probability, to contain the true parameter value. A 95% confidence interval, for example, implies that were the estimation process repeated again and again, then 95% of the calculated intervals would be expected to contain the true parameter value.”
Reason for the name
The term has been derived from the expression "lifting oneself up by one's own bootstraps". This refers to raising oneself up by one's own means. This is also called bootstrap funding.
Example
Let’s understand this with an interested example of a candy factory. Knowing that the candy bars have mean weight. Since it is not feasible to weigh each candy that is produced in the factory, we use sampling techniques and randomly choose 100 candies. We calculate the mean of these 100 candies, and say that the population mean falls under a margin of error.
Suppose that after a couple of months, we want to know the mean candy weight on the day that we sampled the production line (with greater accuracy - or less of a margin of error).
We simply cannot use candies being produced today, as too many elements are now being used like; sugar, different flavours of milk, cocoa beans, different atmospheric conditions, different employees etc. All that we have from the day that we are curious about are the 100 weights.
When to use bootstrapping?
We can simply explain bootstrapping taking simple random sample from the observed sample of same size but with a replacement. This would mean each observation has equal probability of selecting in the sample.
We can use bootstrapping for inference and