Sequence Models for Time Series and Natural Language Processing on Google Cloud training

In this course, we’ll learn how to make predictions on sequences of data.

What you'll learn

In this course, we’ll learn how to make predictions on sequences of data. We’ll cover common business use cases like- 1.time-series prediction and how to deal with more recent data points getting more relevance 2.translating entire sentences (aka sequences of words) into other languages You will get hands-on practice building and optimizing your own text classification and sequence models on a variety of public datasets in the labs we’ll work on together.

Table of contents

Working with Sequences

• Course Introduction
• Sequence data and models
• From sequences to inputs
• Modeling sequences with linear models
• Getting started with GCP and Qwiklabs
• intro:using linear models for sequences
• Time Series Prediction with a Linear Model
• solution:using linear models for sequences
• Modeling sequences with DNNs
• intro:using DNNs for sequences
• Time Series Prediction with a DNN Model
• solution:using DNNs for sequences
• Modeling sequences with CNNs
• intro:using CNNs for sequences
• Time Series Prediction with a CNN Model
• solution:using CNNs for sequences
• The variable-length problem

Recurrent Neural Networks

• Introducing Recurrent Neural Networks
• How RNNs represent the past
• The limits of what RNNs can represent
• The vanishing gradient problem

Dealing with Longer Sequences

• Introduction
• LSTMs and GRUs
• RNNs in TensorFlow
• Intro: Time series prediction:end-to-end (rnn)
• Time Series Prediction with a RNN Model
• Solution: Time series prediction:end-to-end (rnn)
• Deep RNNs
• Intro: Time series prediction:end-to-end (rnn2)
• Time Series Prediction with a Two-Layer RNN Model
• Solution: Time series prediction:end-to-end (rnn2)
• Improving our Loss Function
• Demo: Time series prediction:end-to-end (rnnN)
• Working with Real Data
• Intro: Time Series Prediction - Temperature from Weather Data
• An RNN Model for Temperature Data
• Solution: Time Series Prediction-Temperature from Weather Data
• Summary

Text Classification

• Working with Text
• Text Classification
• Selecting a Model
• Intro: Text Classification
• Text Classification using TensorFlow/Keras on AI Platform
• Solution: Text Classification
• Python vs Native TensorFlow
• Text Classification with Native TensorFlow
• Summary

Reusable Embeddings

• Historical methods of making word embeddings
• Modern methods of making word embeddings
• Introducing TensorFlow Hub
• Intro: Evaluating a pre-trained embedding from TensorFlow Hub
• Using pre-trained embeddings with TensorFlow Hub
• Solution: TensorFlow Hub
• Using TensorFlow Hub within an estimator

Encoder-Decoder Models

• Introducing Encoder-Decoder Networks
• Attention Networks
• Training Encoder-Decoder Models with TensorFlow
• Introducing Tensor2Tensor
• Intro: Cloud poetry:Training custom text models on Cloud ML Engine
• Text generation using tensor2tensor on Cloud AI Platform
• Solution: Cloud poetry:Training custom text models on Cloud ML Engine
• AutoML Translation
• Dialogflow
• Intro: Introducing Dialogflow
• Getting Started with Dialogflow
• Solution: Dialogflow

Summary

contact us

email - [email protected]