mlatoz

Classification Model Selection in Python

Confusion Matrix & Accuracy Ratios

Accuracy Rate and Error Rate:

    AR = Correct/Total = (TN + TP)/Total = 84/100 = 84%
    ER = Incorrect/Total = (FN + FP)/Total = 16/100 = 16%

where,
    AR = Accuracy Rate
    TN = True Negative
    TP = True Positive
    
    ER = Error Rate
    FN = False Negative
    FP = False Positive

---

Additional Reading

Understanding the Confusion Matrix from Scikit learn

Samarth Agrawal (2021)

Link: https://towardsdatascience.com/understanding-the-confusion-matrix-from-scikit-learn-c51d88929c79

---

Comparison between Accuracy Scores of different Classification Models

Classification Model Name	Accuracy Scores
Logistic Regression	94.73
K-Nearest Neighbors	94.73
Support Vector Machine	94.15
Kernel SVM	95.32
Naive Bayes	94.15
Decision Tree Classification	95.90
Random Forest Classification	93.56

Note:- Here the highest accuracy score is of Decision Tree Classification

---

Download Resources

• Regression Files | ZIP
• Classification Files | ZIP
• Model Selection | ZIP

---

«Previous

Next»