Write two functions; One should return the distance measures using Euclidean distance and another one should use mahalanobis distance measure.
Hi @wehired you can use scipy’s functions scipy.spatial.distance.euclidean( ) andscipy.spatial.distance.mahalanobis( ) to calculate Euclidean and Mahalanobis distance, respectively.
Do you want to write the code from scratch?
You can use the predefined functions as well to calculate these distances.
from scipy.spatial import distance dst = distance.euclidean(a, b) # where a and b are two arrays
from scipy.spatial import distance dst = distance.mahalanobis(a,b,VI) # here a and b are arrays and VI is the inverse of covariance matrix
Refer here for more details.
Thank you very much to give euclidean and Mahalanobis functions but as a beginner am still finding it hard to apply it below; please let some one show me how.
Formulate a real world classification task/problem- whose dataset can be applied on a KNN algorithm to design a machine learning model that can be used to solve the task/problem in future. This part includes various activities such as:
- Generating the task be a binary label
- Generating the dataset (a minimum of 50 records), this can be a binary labelled data set or multi-class dataset.
- Separate the features from the labels using the slice method of lists, split the dataset into two: train and test using the train-test split on a ratio of 0.2
- Using the sklearn documentation on KNN, train the algorithm on the KNN classifier and the train dataset.
- Run a predict function on the test set.