Deep Learning

Table of Contents


Deep learning is incredibly powerful tool for extracting complex pattern from data using neural network having multiple layer in it.

Following figure depicts what deep learning is and where it resides in the field of artificial intelligence:

Deep Learning Introduction
Figure: Deep Learning Introduction

Deep Learning is sub field of Artificial Neural Network which in turn sub field of Machine Learning in Artificial Intelligence.

Artificial Intelligenec: Any technique that mimics human behavior using computer or digital processor is known as artificial intelligence. For example: Robots, Chatbots, Spam Filter & Email Categorization, Face Recognition etc

Machine Learning: Ability to learn by machine from examples or without being explicitly programmed is known as machine learning. For Example: Face Recognition, Stock Market Prediction, Voice Recognition etc

Artificial Neural Network: Computational algorithm for machine learning inspired by human brain is known as artificial neural network. For example: Stock Market Prediction, Object Detection, Face Recognition, Outlier Detection etc.


Deep learning has following applications:

  1. Computer Vision: Object Detection, Object Recognition, Face Recognition etc.
  2. Natural Language Processing: Speech Recognition, Language Understanding, Language Generation etc.
  3. Bioinformatics: Understanding Biological Data, Finding Pattern in Biological Data etc.
  4. Machine Translation: Translating text and speech from one language to another language.
  5. Medical Image Analysis: Analyzing different images in medical field like Medical Resonance Imaging (MRI). Application includes brain tumor detection, cancer diagnosis and detection.


Deep learning has following requirements:

  1. Large Data Requirements: Deep learning algorithms are hungry for huge amount of data to succeed.
  2. Hardware Requirements: Deep learning algorithms are tremendously parallelizable which means they can benefit from modern Graphical Processing Unit (GPU).
  3. Software Requirements: Due to open source toolboxes like Tensorflow, Kaggle, Keras building and developing deep learning algorithms has become extremely streamlined.

These article series are focused on different aspects of deep learning with practical examples implemented in python programming language, Tensorflow & Google Colab.

Google Colab

Building Blocks