1.4 DEEP LEARNING APPLICATIONS

 An introduced earlier, there is an excess of scenarios and applications where Deep Leaning is being used. Let us look at few applications in Deep Learning for a more profound understanding of where exactly DL is applied

1.3 WHAT IS THE NEED OF A TRANSTITION FROM MACHINE LEARNING TO DEEP LEARNING?

 Machine Learning has been around for a very long time. Machine Learning helped and motivated scientists and researchers to come up with newer algorithms to meet the expectations of technology enthusiasts. The major limitation of Machine Learning lies in the explicit human intervention for the extraction of features in the data that we work (Figure 1.1). Deep Learning allows for automated feature extraction and learning of the model adapting all by itself to the dynamism of data.

Apple => Menual feature extraction => Learning => Machine learning => Apple

Limitation fo Machine Learning.

Apple => Automatic feature extraction and learning => Deep learning => Apple

Advantages of Deep Learning.

Deep Learning very closely tries to imitate the structure and pattern of biological neurons. This single concept, which makes it more complex, still helps to come out with effective predictions. Human intelligence is supposed to be the best of all types of intelligence in the universe. Researchers are still striving to understand the  complexity of how the Human intelligence is supposed to be the best of all types of intelligence in the universe. Researchers are still striving to understand the complexity of how the human brain works. The Deep Learning module acts like a black box, which takes inputs, does the processing in the black box, and gives the desired output. It helps us, with the help of GPUs and TPUs, to work with complex algorithms at a faster pace. The model developed could be reused for similar futuristic applications.

1.2 THE NEED: WHY DEEP LEARING?

 Deep Learning application have become an indispensable part of contemporary life. Whether we acknowledge it or not, there is no single day in which we do not use our virtual assistants like Google Home, Alexa, Siri and Cortana at home. We could commonly see our parents use Google Voice Search for getting the search results easily without requiring the effort of typing. Shopaholics cannot imagine shopping online without the appropriate recommendations scrolling in. We never perceive how intensely Deep Learning has invaded our normal lifestyles. We have automatic cars in the market already, like MG Hector, which can perform according to our communication. We already have hte luxury of smart phones, smart homes, smart electrical applicances and so forth. We invariably are taken to a new status of lifestyle and comfort with the technological advancements that happen in the field of Deep Learning.

1.1 INTRODUCTION

 Artificial Intelligence and Machine Learning have been buzz words for more than a decade now, which makes the machine an artificially intelligent one. The computational speed and enormous amounts of data have stimulated academics to deep dive and unleash the tremendous research  potential that lies within. Even though Machine Learning helped us start learning intricate and robust systems. Deep Learning has curiously entered as a subset for AI, producing incredible results and outputs in the field.

Deep Learning architecture is built very similar to the working of a human brain, whereby scientists teach the machine to learn in a way that humans learn. This definitely is a tedious and challenging task, as the working of the human brain itself is a complex phenomenon. Our research in the field has resulted in valuable outcomes to makes things easily understandable for scholar and scientists to build worthy applications for the welfare of society. They have made the various layers in neural nets in Deep Learning auto-adapt and learn according to the volume of datasets and complexity of algorithms.

The efficacy of Deep Learning algorithms is in no way comparable to traditional Machine Learning helped industrialists to deal with unsolved problems in a convincing way, opening a wide horizon with ample opportunity. Natual language processing, speech and image recognition, the entertainment sector, online retailing sectors, banking and finance sectors, the automotive industry, chat bots, recommender systems, and voice assistants to self-driving car are some of the major advancements in the field of Deep Learning.

CHAPTER1. Introduction to Deep Learning

 LEARNING OBJECTIVES

After reading through this chapter, the reader will understand the following:

- The need for Deep Learning

- What is the need of transition from Machine Learning to Deep Learning?

- The tools and languages available for Deep Learning

- Further reading

1.5 Deep Learing Framework

 If a workman wants to be good, he must first sharpen his weapon. After learning about the basic knowledge of deep learning, let's pick the tools used to implement deep learning algorithms.

1.4.3 Reinforcement Learning

 Virtual Games. Compared with the real environment, virtual game platforms can both train and test reinforcement learning algorithms and can avoid interference from irrelevant factors while also minimizing the cost of experiments. Currently, commonly used virtual game platforms include OpenAI Gym, OpenAI Universe, OpenAI Roboschool, DeepMind OpenSpiel, and MuJoCo, and commonly used reinforcement learning algorithms include DQN, A3C, A2C, and PPO. In the field of Go, the DeepMind AlphaGo program has surpassed human Go experts. In Dota2 and StarCraft games, the intelligent programs developed by OpenAI and DeepMind have also defeated professional teams under restriction rules.

Robotics. In the real environment, the ocntrol of robots has also made some progress. For example, UC Berkeley Lab has made a lot of progress in the areas of imitation learning, meta learning, and few-shot learning in the field of robotics. Boston Dynamics has made gratifying achievements in robot applications. The robots it manufactures perform well on tasks such as complex terrain walking and multi-agent collaboration (Figure 1-19).

Autonomous driving is considered as an application direction of reinforcement learning in the short term. Many companies have invested a lot of resources in autonomous driving, such as Baidu, Uber, and Google. Apollo from Baidu has begun trial operations in Beijing, Xiong'an, Wuhan, and other places. Figure 1-20 shows Baidu's self-driving car Apollo.