Entering into the world of Artificial Intelligence! (AI,ML and DL)
The first word one normally comes across in this vast field of Artificial Intelligence (AI) is probably Machine Learning (ML). Machine Learning of ML, as in popular culture is a subset of AI that enables us to train the machine with data to enable it to predict or classify the new data we feed it with.
AI and ML are two different terms. ML is only a subset of AI. Artificial Intelligence is a technology using which we can create intelligent systems that can simulate human intelligence. On the other hand, Machine Learning is a sub-field of artificial intelligence, which enables machines to learn from past data or experiences without being explicitly programmed.
I have given a brief about the major differences between AI and ML.
Now, as we move further into the discussion, we come across another term known as Deep Learning or DL. Deep Learning is relatively a new field compared to AI and ML. Deep Learning (DL) is mainly involved in the construction of Neural Networks which is a circuit composed of artificial neurons. This Neural Network is said to mimic the Biological Neural Network in training the Machine.
Today, Deep learning architectures such as Recurrent Neural Networks (RNN)and Convolutional Neural Networks (CNN) have been applied to fields including computer vision, speech recognition, Natural Language Processing (NLP),and medical image analysis.
Deep Learning (DL) is a originally a subset of Machine Learning (ML) which is mostly applied to Larger datasets. Thus, we can summarize that DL is a sub-set of ML and ML is a subset of AI. This is illustrated in the following Image.
DL gained importance in course of time when the data became large and too complex for ML algorithms to solve them. Over a point of time, as data became more and more, the performance of ML models (algorithms) became stagnant (plateau). It was only then Deep Learning became more recognised.
Slide by Andrew Ng, all rights reserved.
Another important property of DL is that, with more data and bigger models, the results get better and better. This is another major reason that DL gets more importance in today’s world.
To get a clear understanding about the difference between ML and DL, let us consider a case in which we have to classify if a person will buy a product or not with features such as age,salary etc..
In training the ML model, we have to manually input the features and their associated weights to the training data.
On the other hand, in training the DL model, the neural network automatically detects the weights (importance) of each feature of the training data and classifies if the customer will buy the product or not.
Another common example is the “Flashlight example”. Suppose we have a flashlight and we train an ML model to turn on the flashlight when somebody says “dark”, the ML model will analyse different phrases said by people and it will search for the word “dark” and turns on the flashlight.
What if someone said “I am not able to see anything the light is very dim”, here the user wants the flashlight to be on but the sentence does not the consist the word “dark” so the flashlight will not be on. That’s where Deep Learning is different from Machine Learning. If it were a deep learning model it would on the flashlight, a DL model would be able to learn from its own method of computing.
To summarize, we can say that Artificial Intelligence is an umbrella term, and Machine Learning and Deep Learning are the subdomains of this field that help in achieving Artificial Intelligence.
Hopefully, I have given the hierarchical description of AI, ML, DL and cleared the basic uncertainty among these terms. Till then, Happy Programming!
