Z-Score (🚶 Logistic Regression) §

$z=\left({\sum }_{i=1}^n{w}_i{x}_i\right)+b$

• $w_i$ represents the weight (= importance) of a feature $x_i$
• $b$ is an extra bias

📖 Example: §

Sentiment classification §

Data §

Calculation §

$\begin{array}{rl}\text{ Suppose }w& =[2.5,-5.0,-1.2,0.5,2.0,0.7]\\ b& =0.1\end{array}$

$\begin{array}{rl}p(+\mid x)=P(Y=1\mid x)& =\sigma (w\cdot x+b)\\ & =\sigma ([2.5,-5.0,-1.2,0.5,2.0,0.7]\cdot [3,2,1,3,0,4.19]+0.1)\\ & =\sigma (.833)\\ & =0.70\end{array}$

🔗 Links

🚶 Logistic Regression