1, evenly distributed, expected to be (a+b)/2, and the variance is the square of (b-a)/12.

2, binomial distribution, expectation is np, variance is npq.

3. Poisson distribution, the expectation is p, and the variance is p.

4. Exponential distribution, with expected value of 1/p and variance of1/(the square of p).

5. Normal distribution, with expectation of U and variance of&; The square of.

6. If X obeys the 0- 1 distribution with the parameter p, then e(x)=p and d(x)=p( 1-p).

Matters needing attention in variance calculation

The covariance matrix calculates the covariance between different dimensions, not between different samples. According to the following 2, each sample has many features, and each feature is a dimension.

According to the formula, the calculation of covariance needs to calculate the mean, that is, calculate the mean by row or column. Covariance matrix is to calculate the covariance between different dimensions, so we should always remember this.