About the data

This is a generic data about Iris flower. Data is a copy of data from sklearn

Whats special about this analysis

Unlike other ML analysis, in this multiple below models were evaluated to find the best model

• Logistic Regression (LR)
• Linear Discriminant Analysis (LDA)
• K-Nearest Neighbors (KNN).
• Classification and Regression Trees (CART).
• Gaussian Naive Bayes (NB).
• Support Vector Machines (SVM).
• XGBClassifier (XGBoost).

Analysis

Data was straight forward so no data engineering methods were applied.

Once above models were applied, Accuracy score and Cross Validation were calculated

• LogReg
  Accuracy: 83.33%.
  Cross Validation Results Mean: 0.941667.
  Cross Validation Results STD: 0.065085.

• LnrDisAnal
  Accuracy: 100.00%.
  Cross Validation Results Mean: 0.975000.
  Cross Validation Results STD: 0.038188.

• KNN
  Accuracy: 100.00%.
  Cross Validation Results Mean: 0.958333.
  Cross Validation Results STD: 0.041667.

• DecTreeClass
  Accuracy: 96.67%.
  Cross Validation Results Mean: 0.933333.
  Cross Validation Results STD: 0.050000.

• GausNB
  Accuracy: 96.67%.
  Cross Validation Results Mean: 0.950000.
  Cross Validation Results STD: 0.055277

• SVM
  Accuracy: 96.67%.
  Cross Validation Results Mean: 0.983333.
  Cross Validation Results STD: 0.033333.

• XGB
  Accuracy: 96.67%.
  Cross Validation Results Mean: 0.958333.
  Cross Validation Results STD: 0.041667.
  Based on above, LDA comes out to be best model to go for prediction.

Visualizing Predicted VS Real Plot using LDA