Let’s first ask ourselves this simple question; what is “deep” in deep learning? What makes deep learning special?
For deep learning to reach full potential, we need to radically democratize it. That is to say, for a future or current practitioner of deep learning, it’s important to be able to recognize the signal in the noise so that you can tell world-changing developments from overhyped press releases.
The “deep” in deep learning isn’t a reference to any kind of deeper understanding achieved by the approach; rather, it stands for the idea of successive layers of representation. Therefore, the number of layers that contribute to the model of data is called depth. Now, this begs the question, what is the anatomy of a neural network?
The network is composed of layers that are chained together and map the input data to predictions. Besides, it also consists of an objective function called the loss function; which is the quantity to minimized during training. The loss function then compares these predictions to the targets, producing a loss value: a measure of how well the network’s predictions match what was expected. There and then, the optimizer uses this loss value to update the network’s parameters called weights!
2. How AI Will Power the Next Wave of Healthcare Innovation?
3. Machine Learning by Using Regression Model
4.Top Data Science Platforms in 2021 Other than Kaggle
What makes deep learning special? There’s a much broader variety of network topologies. Common ones being: Two-branch networks, Multihead networks, Inception blocks among others. These topologies define hypothesis spaces. A hypothesis space is basically useful representations of some input data, within a predefined space of possibilities, using guidance from a feedback signal. Therefore each network has its space of possibilities.
Well, there’s also an optimization technique which is the engine of neural networks called the gradient descent and data representations in neural networks. Very vital. In a bid to democratize deep learning, we are choosing to approach it batch by batch. As you will later realize that’s the only way deep learning models consume input!
More content about this topic will be published sequentially.