script.txt

Problem statement: Quantum generalization of perceptron using neural networks 


1) Objective: Optimizing the algorithm and the loss function to handle the noisy data

ANN was developed in 1950s
ANN + Quantum Computing = QNN
QNN has significant advantage over classical computation.

> As Moore’s law (transistors are doubling every two years in a device)approaches its end, 
two new computing paradigms have been explored, neuromorphic and quantum computers. 

2) NISQ devices.( 'Noisy Intermediate-Scale Quantum'  devices) 
3) State space, bra and ket vectors, what are QNN
4) Hilbert space  | Phi(i) >
5) tensor product (x)
6) l - number of layers , Ro - Tensor, Mi - Total number of perceptrons acting in l-1 hidden layers


> Our algorithm is designed in a way which can handle noisy data or it is robust to noisy data. 
> Faster
> Efficient

7) Limitations: 

> we dont have fault tolerant quantum computers for solving hard problems
> the best-known disadvantage is their “black box” nature. Simply put, you don’t know how or why your NN came up with a 
certain output.
For example, when you put an image of a cat into a neural network and it predicts it to be a car, 
it is very hard to understand what caused it to arrive at this prediction.