Deep Learning

An introduction

What is Deep Learning?

Deep Learning is a subfield of Artificial intelligence and machine learning which is inspired by the structure of a human brain.

Image credits: analyticsvidhya.com

Since we are introducing Deep learning, let us not focus on definition but rather understand what led to the need for Deep learning even though we have ML techniques available to us and why suddenly everyone has started using deep learning.

1. Data: With the advancement of technology we are able to store lots of data and we are able to convert lots of data that was available earlier but was not available in the proper format. Deep learning is a data-hungry technique and it needs lots and lots of data. Machine learning, on the other hand, performs well with minimal data, and as the data grow the result becomes consistent.

Image Credits: Dzone.com

2. Hardware Dependency: ML algorithms can be trained on a normal CPU while Deep Learning models need GPU for better performance. On a normal CPU, these algorithms might not run or will run very slowly because these algorithms are much more complex.

3. Training time: As Deep learning models are complex, and they work with a very large number of data, the training time is very high for these models. In simple words, you can say Deep learning models need more time and data to be sure about the result as compared to machine learning algorithms.

4. Feature Selection: Deep learning has the capability to automatically extract features from the data while in ML we need to manually provide the model with extracted features.

Types of Algorithms in Deep Learning.

1. ANN: Artificial Neural Network
2. CNN: Convolutional Neural Network
3. RNN: Recurrent Neural Network.
4. GAN: Generative Adversarial Network

Why is Deep Learning getting so much attention?

1. Applicability: Deep learning is being used in various domains nowadays. Knowing or unknowing, we use various Deep learning applications on a daily basis. The smartphone has an AI-enabled feature which can easily tell if the image is food or a tree or a human. We can search using speech which internally converts speech to text and then perform a search. There are various games available where we play with machines which is again a deep learning application. Deep learning is used in various domains, and research fields where it gives results like never before.
2. Performance: Deep learning performed much better as compared to machine learning. In many scenarios, Deep learning application has outperformed human results.
3. Library and trained Models: With the advancement of Deep learning techniques now there are various libraries available that can perform complex tasks along with pre-trained models which saves time of training.

Coming back to our first question of what is deep learning?

We can say deep learning is an advanced technique that helps us to solve complex problems using various neural network algorithms.

This is a very short introduction to deep learning. I will try to explain it in more technical detail in the coming days.

You can connect with me on LinkedIn:https://www.linkedin.com/in/aditya-kumar-2a4445b6/