Selamat!

Generalized Linear Models di Python

Ita Cirovic Donev

Data Science Consultant

MODEL

  • Data $\rightarrow$
  • Fungsi taut $\rightarrow$
  • Model $\rightarrow$
  • Kenaikan 1 unit pada $x \rightarrow$

REGRESI LOGISTIK

  • Biner
  • Logit
  • $logit(y) = \beta_0+\beta_1x_1$
  • menaikkan $\color{red}{\text{log odds}}$ sebesar $\beta_1$

MODEL LINIER

  • Kontinu
  • Identitas
  • $y = \beta_0+\beta_1x_1$
  • menaikkan $\color{red}{y}$ sebesar $\beta_1$

REGRESI POISSON

  • Hitung
  • Logaritma
  • $log(\lambda) = \beta_0+\beta_1x_1$
  • $\color{red}{\text{mengalikan } \lambda}$ dengan $exp(\beta_1)$
Generalized Linear Models di Python

FUNGSI PYTHON UTAMA

  • Fitting model
    statmodels $\rightarrow$

MODEL LINIER

glm('y ~ x', data)

glm('y ~ x', data, 
    family = sm.families.Gaussian())

REGRESI LOGISTIK

glm('y ~ x', data, 
    family = sm.families.Binomial())

REGRESI POISSON

glm('y ~ x', data, 
    family = sm.families.Poisson())
Generalized Linear Models di Python

Langkah berikutnya...

  • Kursus DataCamp
  • Buku rujukan unggulan
    • Regression Modeling Strategies oleh Frank E. Harrell, Jr.
    • An Introduction to Categorical Data Analysis oleh Alan Agresti
    • Applied Predictive Modeling oleh Max Kuhn dan Kjell Johnson
Generalized Linear Models di Python

Selamat memodelkan!

Generalized Linear Models di Python

Preparing Video For Download...