Author(s): Tan Pengshi Alvin

Originally published on Towards AI.

Exploring the concept and applications of Graphs, and how to apply Neural Networks to it
Image by Alina Grubnyak on Unsplash

We have seen the power of Machine Learning in drawing insights about data (Unsupervised Learning) or predicting new outcomes (Supervised Learning), by statistically learning about inherent patterns and relational structures within it. Now, apart from the task — Supervised Learning or Unsupervised Learning — the Machine Learning process can also be broken down into two components: models and data types.

Let's talk about models first. Common traditional Machine Learning models include Random Forest, Support Vector Machines (SVM) and XGBoost, and these models do reasonably well when the data is not very large and does not have complex inherent relations. When scale and complexity come into the picture, the SOTA models are typically based on Neural Network (Deep Learning) architectures. A contrast between traditional Machine Learning and Deep Learning can be seen in the other articles I crafted:

Developing deep neural networks from scratch with Mathematics and Python

pub.towardsai.net

Comprehensive layman introduction to the Decision Tree and Random Forest algorithm

towardsdatascience.com

Neural Networks have another strong advantage, though — their architectures are highly flexible and customizable to data types. The traditional Machine Learning models are commonly applied to one-dimensional data, or tabular data — the kind of data that you see along… Read the full blog for free on Medium.

Join thousands of data leaders on the AI newsletter. Join over 80,000 subscribers and keep up to date with the latest developments in AI. From research to projects and ideas. If you are building an AI startup, an AI-related product, or a service, we invite you to consider becoming a sponsor.

Published via Towards AI