Deep Learning with Tensorflow - The Sequential Problem Video Lecture -

Hello, and welcome!
In this video, we will provide an overview
of sequential data, and explain why it poses
a problem for traditional neural networks.
Whenever the points in a dataset are dependent
on the other points, the data is said to be
sequential.
A common example of this is a time series
where each data point represents an observation
at a certain point in time.
There are other examples of sequential data,
like sentences, stock prices, and even gene
sequences.
But traditional neural networks typically
can't handle this type of data.
To see why, let's consider a non-sequential
problem that's well-suited for a basic neural network.
Suppose we have a dataset that contains temperature
and humidity values, and our goal is to build
a traditional neural network that can predict
if the weather is sunny, rainy, or overcast.
This is a straightforward task for the neural
network.
Using our dataset, we first feed a data point
into the input layer.
The data then flows to the hidden layer, where
the weights and biases are applied.
The output layer classifies the results from
the hidden layer, and then produces the final prediction.
It's important to note that the model does
not remember the data that it just analyzed.
All it does is accept input after input, and
produce individual classifications.
A traditional neural network assumes that
the data is non-sequential, and that each
data point is independent of the others.
As a result, the inputs are analyzed in isolation,
which can cause problems if there are dependencies
in the data.
In practice, this is not much different from
random sampling.
To see how this can be a limitation, let's
go back to the weather example.
It could be the case that the prior day's
weather has some influence on the weather
in the following days.
So for example, if it's sunny today, it
may be more likely to be sunny tomorrow.
A traditional neural network model could not
use this information, so we'd have to turn
to a different type of model, like a recurrent
net.
By now, you should understand why sequential
data cannot be properly analyzed by a traditional
neural network.
Thank you for watching this video.