Deep learning
ML is not just about forecasting numbers but also solving complex problems using neural language processing. These use cases include complex scenarios processed by the human brain, such as building an automated chatbot impersonating humans, reading handwritten text, image recognition, transcribing videos/audios, and converting text to audio and vice versa. Deep learning has the ability to solve such use cases by mimicking the human brain.
While ML needs a pre-defined set of labeled data using supervised learning, deep learning uses a neural network for unsupervised learning to simulate human brain behaviors by using a large amount of data to develop learning capabilities for machines. Deep learning is a neural network of multiple layers where you don't need to do data labeling upfront. However, you can use both labeled data and unlabeled data with deep learning, depending upon your use case. The following diagram shows a simple deep learning model:
