Deep Learning through first principle: 01
Amit jha
Posted on April 24, 2020
Machine learning has truly captured the attention of the big and small alike. The adaption has reached such a state that there is a lack of deep learning experts in the industry. In this scenario, it becomes important for every deep learning practitioners to not just learn to "build" deep learning models but also learn about them through first principle. In this series, we will break down the deep learning development cycle and write them in PyTorch.
But today, let's talk about deep learning in general and try to understand it with an example. Let's go...
Let's try to connect the neural network(NN) with our life. Imagine, you are a neuron in a vast network. The decisions you make are directly or indirectly affected by every other neuron in the network(other people). You may argue that a kid living in a remote place on earth would never affect your life in any way, which is true. But in that case, the effect of that child's actions(or existence) would be infinitely small and therefore nonexistent. However, that does not mean the kid is not a part of this network. We all are.
Similarly, your actions affect everyone's life. The people you know and the people you don't. Interestingly, this concept is not limited to humans. Everything in existence--dead, alive or lifeless--is part of this network. In this way, we can propose that the entire world is nothing but a vast neural matrix that functions as though it is one.
Your action A is the cause of someone's(or something's) influence W and your experience E towards a known or unknown goal G.
In this series, we will be diving into all these variables and discussing in greater detail about what else triggers and affects A, where does E come from, what is the concept of W and most importantly what is G.
Make sure you follow me and save this post because I will be updating the index here as I add newer content.
Also, hit the β€οΈ button if you liked this post
Cheers,
Jha
Posted on April 24, 2020
Join Our Newsletter. No Spam, Only the good stuff.
Sign up to receive the latest update from our blog.
Related
November 17, 2024