- What is Deep Learning and why is it so popular?
- Is Deep Learning just a hype or does it have real-life applications?
- What is the difference between Deep Learning and Machine Learning?
- What are the prerequisites for starting out in Deep Learning?
- Is it necessary to have a research background in Deep Learning to start out in it?
- What is Deep Learning and why is it so popular?
I think the answer is a bit complex. The popularity of DL is due, in my opinion to a combination of some recent and spectacular successes, the availability of GPUs and cheap computation units that allow to run these models and... a fancy name.
- Is Deep Learning just a hype or does it have real-life applications?
In my opinion, both. Probably it has been over-marketed and over-hyped. But also has a lot of real-life applications.
- What is the difference between Deep Learning and Machine Learning?
Deep Learning is in fact a part of ML. A Deep Network is a type Neural Network, with added feature extraction stages in the initial layers. And Neural Networks are themselves a group of ML techniques.
In fact, none of the concepts of DL are new. Probably the underlying scientific concept is based on the theory of human visual perception from David Marr (1945–1980).
- What are the prerequisites for starting out in Deep Learning?
The only requisites are knowledge about data processing with a sufficient statistical and mathematical background, a middle-high end class personal computer... and a ton of free time.
- Is it necessary to have a research background in Deep Learning to start out in it?
Not really. A background in numerical and statistical data analysis I would say is mandatory. Also, a background or basic understanding of Machine Learning and Image Processing is strongly advisable. But of course you can learn all of this.
A common mistake is to jump straight to learn DL. Using DL techniques (or any other data processing technique) is in itself very easy nowadays (almost all of the work will be performed by a software library), but doing it 'well' is not.
Most errors you will do will be due to incorrectly handling/knowing the data you give to the software, or incorrectly interpreting the results the software gives back to you. There sre a lot of nuances here, and this is the reason why almost all the people who jumps to DL from other fields, fail.
- What is Deep Learning and why is it so popular?
I think the answer is a bit complex. The popularity of DL is due, in my opinion to a combination of some recent and spectacular successes, the availability of GPUs and cheap computation units that allow to run these models and... a fancy name.
- Is Deep Learning just a hype or does it have real-life applications?
In my opinion, both. Probably it has been over-marketed and over-hyped. But also has a lot of real-life applications.
- What is the difference between Deep Learning and Machine Learning?
Deep Learning is in fact a part of ML. A Deep Network is a type Neural Network, with added feature extraction stages in the initial layers. And Neural Networks are themselves a group of ML techniques.
In fact, none of the concepts of DL are new. Probably the underlying scientific concept is based on the theory of human visual perception from David Marr (1945–1980).
- What are the prerequisites for starting out in Deep Learning?
The only requisites are knowledge about data processing with a sufficient statistical and mathematical background, a middle-high end class personal computer... and a ton of free time.
- Is it necessary to have a research background in Deep Learning to start out in it?
Not really. A background in numerical and statistical data analysis I would say is mandatory. Also, a background or basic understanding of Machine Learning and Image Processing is strongly advisable. But of course you can learn all of this.
A common mistake is to jump straight to learn DL. Using DL techniques (or any other data processing technique) is in itself very easy nowadays (almost all of the work will be performed by a software library), but doing it 'well' is not.
Most errors you will do will be due to incorrectly handling/knowing the data you give to the software, or incorrectly interpreting the results the software gives back to you. There sre a lot of nuances here, and this is the reason why almost all the people who jumps to DL from other fields, fail.
Deep Learning is become a revolution of the 21st century. You’ll find a large number of students, techies, professionals, etc really interested in learning Deep Learning.
In order to learn Deep Learning, you need have knowledge on any one programming language, preferably - Java, Python, C++. Apart from this, you also need to have a clear understanding of Machine Learning and its concepts because Deep Learning is a subset of Machine Learning. Having knowledge on few Machine Learning algorithms will be very helpful, for example: Linear Regression, Logistic Regression, K-Nearest Neighbors, K-Means Clustering, etc.
To have a deeper and meaningful understanding of how Deep Learning models works, you got to have good understanding of mathematical concepts like Linear Algebra, Statistics, Probability and a bit of geometry as well.
After you meet all these prerequisites, you can directly jump into understanding what is Deep Learning, how is it related to human neural network, what is an artificial neural network and get a clear picture of the applications of Deep Learning and the fields in which it is being used.
You can then learn the mechanisms behind Deep Learning algorithms and how they are implemented. Here you will come conceptual topics like Weights, Biases, Activation Functions, BackPropagation, Learning Rate, Epochs, etc.
To build any Deep Learning model, you should have knowledge and understanding of any one Deep Learning libraries like TensorFlow, Keras, Py-Torch or Theano.
Dear Bhanu PRAKASH Kolla ,
The prerequisites for really understanding deep learning are linear algebra, calculus and statistics, as well as programming and some machine learning. Theprerequisites for applying it are just learning how to deploy a model.
https://en.wikipedia.org/wiki/Deep_learning
https://deeptalk.lambdalabs.com/t/deep-learning-prerequisite-subjects/34
Regards,
Shafagat
In task based , problem solving approaches to learning, the cognitive complexity of tasks has a direct bearing on students' levels of processing. As such, increasing the involvement load index of input , teachers control the level of processing . It is generally agreed today that the higher the involvement load, the deeper the level of learning.
Deep Learning is become a revolution of the 21st century. You’ll find a large number of students, techies, professionals, etc really interested in learning Deep Learning.
In order to learn Deep Learning, you need have knowledge on any one programming language, preferably - Java, Python, C++. Apart from this, you also need to have a clear understanding of Machine Learning and its concepts because Deep Learning is a subset of Machine Learning. Having knowledge on few Machine Learning algorithms will be very helpfu
- Badges
- Science topic