Download as pdf or txt
Download as pdf or txt
You are on page 1of 1

Management of V.U.C.A.

(Volatility, Uncertainty, Complexity … 11

Fig. 4 Categorization of machine learning algorithms

Broadly, there are three machine learning algorithms, such as supervised, unsu-
pervised and reinforcement learning (Fig. 4).
In supervised machine learning both input and desired output are provided to the
machine before train the dataset. Then minimize the error calculated by comparing
the calculated outputs and the desired output. Some examples of supervised learning
algorithms are Logistic regression, Linear regression, K-Nearest Neighbor (KNN),
Support Vector Machine (SVM), Naïve Bayes, Neural Network etc. It is categorized
by two sub categories, such as classification and regression algorithms. Classification
algorithms classify the problem to produce a categorical solution and the output is
having defined or discrete values whereas regression algorithms are used to produce
continuous values in output.
In unsupervised learning, only input data are fed through the algorithm. No previ-
ously desired outputs are provided. Algorithms adopt features of their own to train the
dataset. It is categorized by two sub categories, such as clustering and data compres-
sion algorithms. In clustering process, the data are grouped based on their similarities
and dissimilarities. E.g. K-Means Clustering, Hierarchical Clustering, Probabilistic
Clustering etc. where as in data compression process redundant data are minimized
and make the minimized data efficient as it’s necessity. E.g. Dimensionality reduction
process (PCA).
Reinforced learning is a type of machine learning which learns of their own
actions and experience and accordingly interact with the interactive environment.
It’s all about taking suitable action for a particular situation which is resulted in
receiving an award. E.g. Markov Decision Process.

You might also like