LSTM layer in TensorFlow - Cheat Sheet & Quick Revision

Recall & Review

beginner

What does LSTM stand for in machine learning?

LSTM stands for Long Short-Term Memory. It is a type of neural network layer designed to remember information for long periods.

beginner

How does an LSTM layer help with sequence data?

An LSTM layer keeps track of important information over time by using gates to control what to remember, forget, or output. This helps it learn patterns in sequences like sentences or time series.

intermediate

Name the three main gates inside an LSTM cell.

The three main gates are: Forget gate (decides what to forget), Input gate (decides what new information to add), and Output gate (decides what to output).

intermediate

What is the purpose of the cell state in an LSTM layer?

The cell state acts like a conveyor belt that carries important information through the sequence, allowing the LSTM to keep or discard information as needed.

beginner

In TensorFlow, how do you add an LSTM layer to a model?

You can add an LSTM layer using: tf.keras.layers.LSTM(units, return_sequences=True/False). 'units' is the number of memory cells.

What problem does an LSTM layer mainly solve?

A Reducing model size

B Remembering long-term dependencies in sequences

C Improving image resolution

D Sorting data faster

LSTM layers are designed to remember important information over long sequences, which helps with tasks like language or time series.

Which gate in an LSTM decides what information to forget?

A Forget gate

B Input gate

C Output gate

D Reset gate

The forget gate controls which information is removed from the cell state.

In TensorFlow, which argument controls if an LSTM layer returns the full sequence or just the last output?

A return_sequences

B activation

C input_shape

D units

The 'return_sequences' argument decides if the LSTM outputs the full sequence or only the last output.

What type of data is an LSTM layer best suited for?

A Images

B Tabular data

C Sequential data like text or time series

D Static numbers

LSTMs are designed to handle sequential data where order and context matter.

Which of these is NOT a part of an LSTM cell?

A Output gate

B Forget gate

C Input gate

D Pooling layer

Pooling layers are used in convolutional networks, not inside LSTM cells.

Explain how an LSTM layer processes sequence data and why it is better than a simple RNN.

Think about how LSTM remembers important info over time.

Describe how to add an LSTM layer in TensorFlow and what key parameters you need to set.

Focus on the code and what controls output shape.