Clustering

This dataset has 12 columns and 217 rows.
The preview is meant for you to glance at the data and understand it.
This view shows all columns and the first 100 rows.

Choose one or more of these pre-defined feature groups to cluster by.

	Name	Description
	Absolute marks	Cluster based on absolute marks Maths, Science, General, Essay, Average
	Percentile marks	Cluster based on marks percentiles Math PC, Science PC, General PC, Essay PC, Average

Choose individual features you want to cluster by. You should pick the features that will have the highest impact on your problem area. You can select between 2 – 40 features.

Please correct these

Maths

Performance in Mathematics

Science

Performance in Science

General

Performance in General

Essay

Performance in Essays

Math PC

Mathematics percentile score

Science PC

Science percentile score

General PC

General percentile score

Essay PC

Essay percentile score

Average

Average performance across subjects

Choose the algorithm you want to use for clustering. If you're unsure, try them one by one.

K-Means

Group points districts closest to each other on features

Mini Batch K-Means

Same as K-Means, but faster and slightly worse

Spectral Clustering

Same as K-Means, but reduces number of features first

Affinity Propagation

Find most representative entities and group nearby ones

Expectation Maximization

Similar to K-Means, but doesn't assign hard clusters to points

Mean-Shift

Find smooth blobs

Ward

Build clusters one-by-one using hierarchically

Density Based Scan

Find dense regions (of any shape) seperated by sparse regions

HDBSCAN

Same as HBSCAN, but density of regions can vary

To override the default number of clusters, pick a value. (Leave it blank for the optimal default.)

Number of clusters:

For help or information regarding the application, please email contact@gramener.com.

Student Marks

Please correct these