Feb 01, 2019 19:43:32

# Writing is revising

great writers revise their writings numerous times and they love revising and think of audience.

It was an eye-opening for me since I thought those who are good at writing can write as breathing, and just produce a better quality of writings instantly. I determined I was not a good writer just looking at the first version.

Easy to come up with the reason not to do so. Infinite reasons exist and assure us we do not have to do this because of them.

Today let me try to revise the first version of the post "FT1 - Bayes Theorem." Not fool myself and not just make a similar post, I asked feedback from my colleague and discuss the contents to come up with the new version as below.

### Bayes theorem

Bayes theorem helps us to improve and update the prior knowledge of some events based on other observation or findings. We are using this in daily life. For example, if it rains today, you will think it will rain tomorrow high likely in a rainy season.

Here is another example for a kid with using numbers.

Let's say a boy, named John, would like to know how his mom, Mary's,  mood was.

• Her mood can be only in a "good" or "bad" mood.
• Her activity can be only 'staying home' or 'going out.'

Here are other conditions:

• John also knows overall 80% of the time Mary is in a good mood.
• If she was in a 'good' mood, she was also "going out" with 90% probability
• If she was in a 'bad' mood, she was likely 'staying home' with 70% probability.

So today John is coming back from school and ask if Mary was 'going out' or 'staying home.' Her answer was 'going out.' Based on this new information, John will update the general knowledge "she is normally 80% in a good mood" as follows:

• "going out" and 'good mood' can happen 0.8(80%) x 0.9(90%) = 0.72
• "going out" in spite of being in 'bad mood' can happen 0.2(20%) x 0.3(30%) = 0.03
• The ratio of "going out" and "good mood" over "going out" regardless of her mood is 0.72 over 0.72 + 0.06, which is 72/78 almost 0.92 (92%)

In conclusion, John can update the probability from 80% to 92% based on his new observation today.

contact: email - twitter / Terms / Privacy