SHY.John Place

Had some ideas to continue my Python statistics tutorial with adding the two datasets:

  • A hypothetical bank loan data for logistic regression
  • A hypothetical survey data that mimics the Kaggle survey.

Added them on the GitLab and will add the tutorial instructions soon… If not this weekend.

WordPress is asking me to upgrade my PHP version. Just wondering if that will break my blog…

Realised I’ve left this place for so long…

Originally I thought I wouldn’t come back to this place. But after getting a job in tech field. I realised that I could make this as my portfolio.

貝氏概率 (Bayesian probability)

貝氏概率是指我們計算一件事出現的概率時，除了藉由它們出現的頻率來計出外我們有時會用機率分佈來定出一件事出現的概率。機率分佈是一種函數，即便它一定會有參數來顯示它的形狀和位置。貝氏概率將這些參數變成機率分佈。即便機率分佈本身包含其他機率分佈。 貝氏概率在主流物理學上應用不多，但之在AI或計算模型分析便很有用。例如我們如果知道實驗室在不同時距會有不同壞掉的機率，這是一種機率分佈，我們如果知道這個機率分佈是由最有可能壞掉的時距來決定的，然後我們又知道這個參數是由另一個機率分佈來決定（參數可以是當時室溫、濕度等等），是貝氏概率的應用。