Loading...
Deep Learning
"A Mathematical Function Maps Inputs to Outputs using Approximations. Deep Learning is our Best Known 'Universal' Approximator. Its the 'Best' Option we have had so far in a long time." - Founders, Automatski
"But, in the end its just an Approximation, it has Errors, and it doesn't have any Sense or Intelligence, or any Feelings or Thoughts, it cannot Learn on its own and is a Pain, Cumbersome & Expensive to Train. A Layman and most People don't understand that." - Founders, Automatski
"And it takes Tonnes of Data (Inputs and Outputs) to Train one. The Human Brain doesn't need all that effort to learn and do what it does. Obviously we have 'Missed' the Target. But Deep Learning has solved a lot of problems nevertheless." - Founders, Automatski
"But, in the end its just an Approximation, it has Errors, and it doesn't have any Sense or Intelligence, or any Feelings or Thoughts, it cannot Learn on its own and is a Pain, Cumbersome & Expensive to Train. A Layman and most People don't understand that." - Founders, Automatski
"And it takes Tonnes of Data (Inputs and Outputs) to Train one. The Human Brain doesn't need all that effort to learn and do what it does. Obviously we have 'Missed' the Target. But Deep Learning has solved a lot of problems nevertheless." - Founders, Automatski
Restricted Boltzmann Machine (RBM)
Deep Belief Networks (DBN)
Convolution Neural Networks (CNNs)
Recurrent Neural Networks (RNNs)
Recursive Neural Tensor Networks (RNTNs)
Long Short-Term Memory (LTSM)
Generative Adversarial Networks(GANs)
Segmentation Networks
(Variational) Auto Encoders
U-Net
Residual Neural Networks (Res-Net)
The Revolution...
"At Automatski, we have Revolutionized Deep Learning by creating the 'Universal' 'Turing Complete' Deep Learning Algorithm based on An Universal Approximation Function. (The - One Human Brain One Universal Algorithm For Everything Hypothesis)." - Founders, Automatski