The black dot is the new instance we are going to predict with K nearest Neighbour Using Euclidean Distance, calculate its distances to its 3 nearest neighbors What is the shape of this dot if k 3 What would be its shapes if k 5 10 points Download the dataset for this assignment at, https drive google com file d 1HCvBMHk1Xc1deEFJF3TrT7GOEmAACbUN view usp sharing This is the dataset of 100 patients' records with their diagnostic result The variable 'diagnosis result' is the target variable, malignant (M) or benign (B) Check the number of null values Drop the column 'id' Replace the target variable with numerical values, 1 for malignant, 0 for benign 10 points Use boxplot and ANOVA test to check the relationship between the features and the target variable Drop the columns that show no significant difference between the group 'M' and group 'B' 10 points Use correlation analysis to drop the columns that have strong correlation relationship with the others (Note If two features have strong correlation with each other, should you drop both or drop one of these two ) 10 points Explain in your own words why feature scaling is important to the KNN algorithm Normalize all the features 10 points Build a K Nearest Neighbor model with k 3 Use a 10 folds cross validation to obtain the average accuracy score for a KNN model when k 3 10 points Set up k values to odd numbers from 1 to 15 Explain why k shall not be even numbers in this case 10 points Build KNN prediction models with the preset k values Use cross validation to obtain the average accuracy scores for the KNN models Find the optimal k 20 points Besides accuracy, there are other metrics to evaluate a classification model, which is especially helpful when the data is imbalanced Learn about precision, recall, and f1 score with the materials below Which scoring metric fits best for this prediction problem Change the scoring option for cross validation to f1 Did the optimal k value change 10 points Precision, Recall and F1 Score, https youtu be sJR 1yz7mnI Precision, Recall and Predicting Cervical Cancer with Machine Learning, https proclusacademy com blog explainer precision recall f1 score classification models

Question

The black dot is the new instance we are going to predict with K nearest Neighbour  Using Euclidean Distance, calculate its distances to its 3 nearest neighbors  What is the shape of this dot if k 3  What would be its shapes if k   5  10 points Download the dataset for this assignment at, https   drive google com file d 1HCvBMHk1Xc1deEFJF3TrT7GOEmAACbUN view usp sharing   This is the dataset of 100 patients' records with their diagnostic result  The variable 'diagnosis result' is the target variable, malignant (M) or benign (B)  Check the number of null values  Drop the column 'id'  Replace the target variable with numerical values, 1 for malignant, 0 for benign  10 points Use boxplot and ANOVA test to check the relationship between the features and the target variable  Drop the columns that show no significant difference between the group 'M' and group 'B'  10 points Use correlation analysis to drop the columns that have strong correlation relationship with the others  (Note  If two features have strong correlation with each other, should you drop both or drop one of these two  ) 10 points Explain in your own words why feature scaling is important to the KNN algorithm  Normalize all the features  10 points Build a K Nearest Neighbor model with k   3  Use a 10 folds cross validation to obtain the average accuracy score for a KNN model when k 3  10 points Set up k values to odd numbers from 1 to 15  Explain why k shall not be even numbers in this case  10 points Build KNN prediction models with the preset k values  Use cross validation to obtain the average accuracy scores for the KNN models  Find the optimal k  20 points Besides accuracy, there are other metrics to evaluate a classification model, which is especially helpful when the data is imbalanced  Learn about precision, recall, and f1 score with the materials below  Which scoring metric fits best for this prediction problem  Change the scoring option for cross validation to f1  Did the optimal k value change  10 points Precision, Recall and F1 Score, https   youtu be sJR 1yz7mnI Precision, Recall and Predicting Cervical Cancer with Machine Learning, https   proclusacademy com blog explainer precision recall f1 score classification models

SolutionInn · Accepted Answer

The Answer is in the image, click to view ...

Question: The black dot is the new instance we are going to predict with K-nearest Neighbour. Using Euclidean Distance, calculate its distances to its 3 nearest

Step by Step Solution