Scribd
Upload
12 views100 pages

OCI AI

Uploaded by

asbalyan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
12 views100 pages

OCI AI

Uploaded by

asbalyan
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 100

RNN LSTM Transformer

Recurrent Neural Network remembers the past and its decisions are influenced by what
it has learned from the past.
Backpropagation Through Time(BPTT)
Long Short-Term Memory(LSTM)
LSTM is an improved version of the regular RNN which was designed to make it easy to
capture long-term dependencies in sequence data. A regular RNN functions in such a
way that the hidden state activation is influenced by the other local activations nearest
to them, which corresponds to a “short-term memory”, while the network weights are
influenced by the computations that take place over entire long sequences, which
corresponds to “long-term memory”. Hence the RNN was redesigned so that it has an
activation state that can also act as weights and preserve information over long
distances, hence the name “Long Short-Term Memory”.

https://www.theaidream.com/post/introduction-to-rnn-and-lstm
Machine Learning Foundation
Multi class classification

Deep Learning
Deep learning models Sequence Models
CNN Convlution Neural Netwoks
FNN Feedforword Neural network
– Also called Multi Layer Perceptron (MLP)
– Simplest form of neural networks

CNN Convolution Neural Network


– Automatically learn local design patterns and features in images, etc

RNN Recurrent Neural Networks


– Designed to handle sequential data, such as time series data or natural language
– Have feedback loop that allows them to maintain hidden states and capture temporal
dependencies

Recurrent Neural Networks (RNNs) are a type of neural network architecture that
includes feedback connections. These feedback connections allow RNNs to process
sequential data such as time series, natural language, speech, and more.
Many-to-Many RNN
Many-to-Many

Machine Translation involves translating a sentence or a sequence of text from one


language to another, which is essentially a sequence-to-sequence problem. In the
Many-to-Many RNN architecture, the network takes a sequence of inputs and produces
a sequence of outputs. In the context of machine translation, this means it can take a
sequence of words or tokens in one language as input and generate a corresponding
sequence of words or tokens in another language as output.

Autoencoders
– These are unsupervised learning model used for feature extraction and dimensionality
reduction, and is commonly employed in data compression and anomaly detection

LSTM Long Short-Term memory


– specialized RNN variant designed to handle long term dependencies in sequential data
Long Short-Term Memory (LSTM) Neural Networks are the sequence model of choice
when it comes to handling and maintaining relevant information over long sequences,
making them particularly well-suited for tasks such as language modeling, machine
translation, and speech recognition.

GAN Generative adversarial Networks


– powerful deep learning model, which is used for generating realistic synthetic data like
images, text etc

Transformers
– Widely used in natural language processing and have become state of the art model for
machine translation, text generation and language understanding.
ANN for single dimensional data and CNN for doule dimensional data
Generative AI and LLM
Neural Probabilistic Language Models
Transformer LLM
Vanishing Radiant Gradiant
Prompt
Lifecycle
Finetuning
OCI ML services
Trust ethical AI
OCI Generative AI service

You might also like