Skip to main content
Data Science

Return to Answer

added 195 characters in body
Source Link
Kari
Kari
  • 2.8k
  • 2
  • 21
  • 50

My question is does LSTM by nature, also take previous output values besides feature vector as new feature element for the next output values?

You didn't specify which framework you are using, so I will explain for a custom (c++ etc) implementation.

At each timestep LSTM will:

  1. Take a new input from you (for example, as a one-hot encoded vector)
  2. Internally fetch a separate vector which is the LSTM's output from Timestep [t-1]
  3. These two vectors are like a 'fuel' and will be used to shape out a Cell at [t]
  4. The result gets produced for this Timestep [t]

Back Propagation creates gradients for the weights within the LSTM, but also gradients for these two vectors.

Because there are multiple timesteps, you get a Gradient during each Timestep, which has to be addedadded (component-wise) before being used.

Note, this last url link has a blue arrow at the bottom right of the screen, so you can click to see the next slides.

My question is does LSTM by nature, also take previous output values besides feature vector as new feature element for the next output values?

You didn't specify which framework you are using, so I will explain for a custom (c++ etc) implementation.

At each timestep LSTM will:

  1. Take a new input from you (for example, as a one-hot encoded vector)
  2. Internally fetch a separate vector which is the LSTM's output from Timestep [t-1]
  3. These two vectors are like a 'fuel' and will be used to shape out a Cell at [t]
  4. The result gets produced for this Timestep [t]

Back Propagation creates gradients for the weights within the LSTM, but also gradients for these two vectors.

Because there are multiple timesteps, you get a Gradient during each Timestep, which has to be added (component-wise) before being used.

My question is does LSTM by nature, also take previous output values besides feature vector as new feature element for the next output values?

You didn't specify which framework you are using, so I will explain for a custom (c++ etc) implementation.

At each timestep LSTM will:

  1. Take a new input from you (for example, as a one-hot encoded vector)
  2. Internally fetch a separate vector which is the LSTM's output from Timestep [t-1]
  3. These two vectors are like a 'fuel' and will be used to shape out a Cell at [t]
  4. The result gets produced for this Timestep [t]

Back Propagation creates gradients for the weights within the LSTM, but also gradients for these two vectors.

Because there are multiple timesteps, you get a Gradient during each Timestep, which has to be added (component-wise) before being used.

Note, this last url link has a blue arrow at the bottom right of the screen, so you can click to see the next slides.

added 16 characters in body
Source Link
Kari
Kari
  • 2.8k
  • 2
  • 21
  • 50

My question is does LSTM by nature, also take previous output values besides feature vector as new feature element for the next output values?

You didn't specify which framework you are using, so I will explain for a custom (c++ etc) implementation.

At each timestep LSTM will:

  1. Take a new input from you (for example, as a one-hot encoded vector)
  2. TakeInternally fetch a separate vector which is the LSTM's output from Timestep [t-1]
  3. These two vectors are like a 'fuel' and will be used to shape out a Cell at [t]
  4. The result gets produced for this Timestep [t]

Back Propagation creates gradients for the weights within the LSTM, but also gradients for these two vectors.

Because there are multiple timesteps, you get a Gradient during each Timestep, which has to be added (component-wise) before being used.

My question is does LSTM by nature, also take previous output values besides feature vector as new feature element for the next output values?

You didn't specify which framework you are using, so I will explain for a custom (c++ etc) implementation.

LSTM will:

  1. Take a new input (for example, as a one-hot encoded vector)
  2. Take a separate vector which is the LSTM's output from Timestep [t-1]
  3. These two vectors are like a 'fuel' and will be used to shape out a Cell at [t]
  4. The result gets produced for this Timestep [t]

Back Propagation creates gradients for the weights within the LSTM, but also gradients for these two vectors.

Because there are multiple timesteps, you get a Gradient during each Timestep, which has to be added (component-wise) before being used.

My question is does LSTM by nature, also take previous output values besides feature vector as new feature element for the next output values?

You didn't specify which framework you are using, so I will explain for a custom (c++ etc) implementation.

At each timestep LSTM will:

  1. Take a new input from you (for example, as a one-hot encoded vector)
  2. Internally fetch a separate vector which is the LSTM's output from Timestep [t-1]
  3. These two vectors are like a 'fuel' and will be used to shape out a Cell at [t]
  4. The result gets produced for this Timestep [t]

Back Propagation creates gradients for the weights within the LSTM, but also gradients for these two vectors.

Because there are multiple timesteps, you get a Gradient during each Timestep, which has to be added (component-wise) before being used.

Improved Formatting
Source Link
Toros91
Toros91
  • 2.4k
  • 3
  • 16
  • 32

My question is does LSTM by nature, also take previous output values besides feature vector as new feature element for the next output values?

You didn't specify which framework you are using, so I will explain for a custom (c++ etc) implementation.

LSTM will:

  1. takeTake a new input (for example, as a one-hot encoded vector)
  2. takeTake a separate vector which is the LSTM's output from timestepTimestep [t-1]
  3. theseThese two vectors are like a 'fuel' and will be used to shape out a Cell at [t]
  4. theThe result gets produced for this timestepTimestep [t]

back propagationBack Propagation creates gradients for the weights within the LSTM, but also gradients for these two vectors.

Because there are multiple timesteps, you get a gradientGradient during each timestepTimestep, which has to be added (component-wise) before being used.

My question is does LSTM by nature, also take previous output values besides feature vector as new feature element for the next output values?

You didn't specify which framework you are using, so I will explain for a custom (c++ etc) implementation.

LSTM will:

  1. take a new input (for example, as a one-hot encoded vector)
  2. take a separate vector which is the LSTM's output from timestep [t-1]
  3. these two vectors are like a 'fuel' and will be used to shape out a Cell at [t]
  4. the result gets produced for this timestep [t]

back propagation creates gradients for the weights within the LSTM, but also gradients for these two vectors.

Because there are multiple timesteps, you get a gradient during each timestep, which has to be added (component-wise) before being used.

My question is does LSTM by nature, also take previous output values besides feature vector as new feature element for the next output values?

You didn't specify which framework you are using, so I will explain for a custom (c++ etc) implementation.

LSTM will:

  1. Take a new input (for example, as a one-hot encoded vector)
  2. Take a separate vector which is the LSTM's output from Timestep [t-1]
  3. These two vectors are like a 'fuel' and will be used to shape out a Cell at [t]
  4. The result gets produced for this Timestep [t]

Back Propagation creates gradients for the weights within the LSTM, but also gradients for these two vectors.

Because there are multiple timesteps, you get a Gradient during each Timestep, which has to be added (component-wise) before being used.

added 13 characters in body
Source Link
Kari
Kari
  • 2.8k
  • 2
  • 21
  • 50
Source Link
Kari
Kari
  • 2.8k
  • 2
  • 21
  • 50