Back in 2009, deep learning was only an emerging field and only a few people recognized it as a fruitful area of research. But soon it gained momentum and is used today for several applications. Speech recognition, image recognition, finding patterns in a dataset, object classification in photographs, character text generation, self-driving cars and many more. Hence it is important to be familiar with deep learning and its concepts.

Tutorials for Beginners on Deep Learning

1.Deep Learning Simplified

If the complicated terminologies make it difficult for you to learn deep learning, then this tutorial will prove to be a boon for you. This is a simplified tutorial on deep learning and its basic concepts. You will learn about neural networks, deep net, deep belief nets, convolutional neural networks, H2O.ai, and This tutorial will give you a basic understanding of deep learning. You will also learn about different kind of models, why & when to choose each one of them. Then it will provide you hands-on experience of deep learning with different use cases. You will also learn about different platforms where you can build your own deep nets and the different libraries available for deep learning. The tutorial is devoid of any mathematical calculations or coding and is best for anyone looking to get a basic idea about deep learning.

2. Tutorial: Deep Learning

In this tutorial on Deep Learning Yoshua Bengio and Yann Lecun explains the breakthroughs brought by deep learning in recent years. After their in-depth research of 30 years, Yoshua & Yann share the insights on how deep learning has transformed machine learning & AI. In this tutorial, you will learn how deep learning allows computational models composed of multiple processing layers to learn a representation of data. These methods have improved speech recognition, visual object recognition, object detection and domains like genomics. This tutorial will take you through the basics of deep learning, discuss its various applications and what challenges it poses in front of us.

3. Deep Learning with Neural Networks & Tensorflow Introduction

If you have been wondering how a neural network works and why recently there is so much of uproar created by them. In this tutorial on introduction to neural networks, you will learn how neural network is able to create powerful models with huge datasets. Understand the structure of neural networks and how each input layer combines together to generate an output. This is the only first video of the complete tutorial, for TensorFlow Basics watch part 2 of the tutorial. To know how to build a neural network model, continue watching part 3 and so on.

4. Neural Networks for Machine Learning

The main idea behind studying artificial neural networks is to understand the style of parallel computation of neurons and their adaptive connections. In this course by Prof. Geoffrey Hinton taught at the University of Toronto you will learn how neural networks and machine learning can bring a revolution in technology. It includes topics such as perceptrons, back propagation, CNN, RNN, gradient descent, Bayesian optimization of hyperparameters and many more topics. This is one of the best courses available out there on deep learning. If you are a deep learning enthusiast, you can't just afford to miss it.

5. Intro to TensorFlow

One of the most popular machines learning library right now is TensorFlow. Though it was built for conducting machine learning and deep neural network research primarily. But because of its versatility, Tensorflow can be used in a variety of applications. Here in this interesting tutorial on TensorFlow, you will learn to build a handwritten digit image classifier in Python in under 40 lines of code. You will also learn how to generate music in TensorFlow, what is Tensorboard, build a neural network and pros & cons of using TensorFlow over other deep learning libraries. This brief tutorial on TensorFlow is a must watch for any novice in deep learning.

6. Neural Networks

The artificial neural network is capable of learning and they need to be trained. There are basically 3 steps for building a machine learning model â?? build it, train it and test it. Once the model is built it can be trained to become better & better at pattern recognition. In these quick 5 min videos, you will learn to build a neural network, build autoencoders and build a recurrent neural network. The codes for each video is also available in the description on youtube.

Application of Deep Learning

1. Google's Deepmind Explained

It was a historic moment when Google's AlphaGo beat the world champion Lee Sedol in the ancient board game GO. It triggered a new wave of technology advancement when a machine succeeded over a human. Google Deepmind claimed to bring the next generation of AI and aims to develop programs which will be smart enough to take actions on their own. This video explains when and why Deepmind was founded. And what revolution it can bring in AI.

2. 9 Cool Deep Learning Applications

Wondering what are some of the interesting the real-life applications of Deep learning and machine learning? This video showcases the real-life applications of deep learning. You will come across some intriguing applications like toxicity detection for different chemical structure, mitosis detection for large images, sequence generation, how a computer program itself plays pong and many more interesting applications.

3. Deep Learning Program learns to Paint

Artificial neural networks are inspired by the human brain and the aims to study the connection between the neurons. In the above videos, we have seen several applications of deep learning. But Neural art happens to be the most amazing and surprising application of deep learning. In this video, you will learn how to paint using deep learning or re-create famous painting using artificial neural networks. All the user needs to do is provide an input photograph and a target image from which the art style will be learned.

Reinforcement Learning

1.Introduction to Reinforcement Learning with Function Approximation â?? Tutorial

Reinforcement learning is a technique developed by machine learning and research communities for making optimal sequential decision making. This tutorial will provide you with a thorough understanding of the underlying formal problem (Markov decision processes) and its core solution methods, including dynamic programming, Monte Carlo methods, and temporal-difference learning. It is focussed on how these methods are combined with a parametric approximation to find good approximate solutions to problems that are otherwise too large to be addressed at all. The speaker will also take you through the recent developments in function approximation, eligibility traces, and off-policy learning.

2. Deep Reinforcement Terrain Learning

In this video, a combination of deep learning and reinforcement learning is depicted which is thought to be useful in solving many extremely difficult tasks. Google DeepMind built a system that can play Atari games at a superhuman level using deep reinforcement learning. This video shows an interesting use of deep reinforcement learning to teach terrain animals to map their movements and avoid obstacles in the way.