Inspiration and original work by the good folks from Google: Intro to text classification with Keras: automatically tagging Stack Overflow posts

https://github.com/tensorflow/workshops/tree/master/extras/keras-bag-of-words

An open source machine learning framework - https://keras.io/

pip2.7 install --user tensorflow
or
pip3 install --user tensorflow

A high-level API built in to TensorFlow - https://keras.io/

pip2.7 install --user keras
or
pip3 install --user keras

See Google BigQuery steps to follow

See Preprocessing data steps

See Processing input features steps

See Processing output labels steps

See Building the model steps

https://en.wikipedia.org/wiki/Cross_entropy

Each of the comments (on the post) can only belong to one post

An Optimizer is a function the model uses to minimize loss

Adam optimizer

https://arxiv.org/abs/1412.6980

More on optimizers

https://en.wikipedia.org/wiki/Hyperparameter_(machine_learning)

See Metrics

See Generating predictions steps

Changing the vocabulary size of the BOW model (Bag Of Words model) uses

Changing batch size, number of epochs, or the dimensionality of the input layer

Increasing the size of the training dataset

Adding dropout to one of the layers to prevent the model from overfitting - https://keras.io/layers/core/

Original and improvised versions of the notebook, python scripts and data can be found in the folder data-scripts-notebooks. Also, includes a version with the Weights and Biases package implemented.

Contributions are very welcome, please share back with the wider community (and get credited for it)!

Please have a look at the CONTRIBUTING guidelines, also have a read about our licensing policy.

Back to main page (table of contents)