Question

Please discuss the types of classifications that we can use with our data analytics. Also, how can these classifications and techniques be applied when using our Business intelligence tools?

Answer 1

"Classification" is a supervised learning where computer program learns from input data, trains itself and uses this trained data to classify new observations/input.

One more important point in Classification is, it's output has labels defined. For example 0 or 1.

Like, ManU will win the match or lose the match! It predicts the discrete value associated with a class.

Grossly, there are two types:

1. Binary Classification: Model predicts either of the two values. i.e. 0 or 1

2. Multi-Class Classification: Model predicts more than one class

In Data Analytics/Machine Learning, with the help of Statistics there are different types of Classification/Supervised Learning available, broadly categorized as:

- Linear Regression

1. Logistic Regression - One or more dependent variables that define outcome

2. Naive Bayes Classifier - Identifies unrelated features

- Decision Trees

It is tree like structure to decide on which branch the data goes and finally all data is classified. It has branches and nodes as a structure.

- Neural Networks

It contains layers containing Neurons where input layer accepts input data, sonme non linear function is applied in middle layers. Then output layer will come up with output. Note that, number of layers depend on complexity of problem.

You can use Business Intelligent tools to solve classification problems by implementing any of the above Classification Techniques depending on the requirement.

Also, there are few more algorithms to implement Classification but those are more over with Statistics, for example, k-nearest neighbor algorithm. But these are hard to implement into Business Intelligence Tools and accuracy can be compromised.