Question

5. How different k-NN is from k-Means algorithm? How do you pick the optimal k-value? (see Elbow method)

Answer 1

How different K-NN is from K-Means Algorthim ? How do you pick the optimal K- Value ?

They are often confused with each other. The 'K' in K-Means Clustering has nothing to do with the 'K' in KNN algorithm. k-Means Clustering is an unsupervised learning algorithm that is used for clustering whereas KNN is a supervised learning algorithm used for classification. k-NN is a supervised algorithm used for classification. What this means is that we have some labeled data upfront which we provide to the model for it to understand the dynamics within that data i.e. train. It then uses those learnings to make inferences on the unseen data i.e. test. In the case of classification this labeled data is discrete in nature. k-Means is an unsupervised algorithm used for clustering. By unsupervised we mean that we don’t have any labeled data upfront to train the model. Hence the algorithm just relies on the dynamics of the independent features to make inferences on unseen data.

The optimal K value usually found is the square root of N, where N is the total number of samples. Use an error plot or accuracy plot to find the most favorable K value.

Answer 2

How different K-NN is from K-Means Algorthim ? How do you pick the optimal K- Value ?

They are often confused with each other. The 'K' in K-Means Clustering has nothing to do with the 'K' in KNN algorithm. k-Means Clustering is an unsupervised learning algorithm that is used for clustering whereas KNN is a supervised learning algorithm used for classification. k-NN is a supervised algorithm used for classification. What this means is that we have some labeled data upfront which we provide to the model for it to understand the dynamics within that data i.e. train. It then uses those learnings to make inferences on the unseen data i.e. test. In the case of classification this labeled data is discrete in nature. k-Means is an unsupervised algorithm used for clustering. By unsupervised we mean that we don’t have any labeled data upfront to train the model. Hence the algorithm just relies on the dynamics of the independent features to make inferences on unseen data.

The optimal K value usually found is the square root of N, where N is the total number of samples. Use an error plot or accuracy plot to find the most favorable K value.