Estimated death number was gained using linear regression coefficients calculated from 7 days rolling number of infection and death until 2021-07-12. Estimated death number (cubic blue) was higher than 7 days rolling number of death (green triangle). Lower death ratio? Under estimation?

Previoussly 21 days interval made the correlation high between daily infected and dead numbers. There were also slight increase of correlation 14 and 28 days interval. These may make by the difference of the amount of work depend on the day of the week. To solve this 7 days syntropy, 7 day moving average may be useful.

From 2020-01-16 to 2021/07/12 (just the day when the last declaration of a state of emergency ) ,nhk news data was used for this analysis. Daily death number was well correlated with the daily infected number with 21 days interval.

Impact of the leaf width on the prostate VMAT about dosimeto-volmetric and delivering parameters Hisato Nagano Shonan Fujisawa Tokushukai Hp. JPN

Conjugate Distributions series 0. Conjugacy 1. Bernoulli and beta 2. Categorical and Dirichlet 3. Poasson and gamma 4. Univariate normal and normal-scaled inverse gamma 5. Multivariate normal and normal inverse Wishart 6. Without conjugacy

The categorical distribution is a discrete distribution that determines the probability of observing one of K possible outcomes. If K = 2, it may equal to Berunoulli distribution. We will concern about the world having four possible parameters (K = 4),

ML,MAP and Bayes' posterior distribution about Bernoulli distribution.

Bernoulli distribution is the simple distribution of 1 or 0 with probability of p or 1-p, respectively. Sample means from this population distribute normally ( mean=p and variance=p(1-p) are expected). We are sometimes confused with the difference of values and means. Value is 1 or 0, and mean is p as population near equal to p as samples.

author: "Hisato Nagano" date: '2015-10-24' Storm Data documents the occurrence of storms and other significant weather phenomena having sufficient intensity to cause loss of life, injuries, significant property damage, and/or disruption to commerce. From the human health point of view, tornade was the most harmful phenomenon. It occupied 37% of fatality and 65% of injury. As to economic losses, property damage was caused by flood. It took 34% of all cause of property damage and 32% of all economic loss. On the contrary, the crop damage was mainly (28%) caused by a drought. But this was only 3% of economic loss caused by a natural disaster. We should take a heavy warning to Wind and Water!