Check that the Python model.py file has been created
Check for a Tensorflow model in the /tensorflow/trained_model directory
Check for successful run of a gcloud ai-platform training job called flights-*
Check that a trained model has been saved to Google Storage
Machine Learning with TensorFlow
In this lab, you learn how to use Google Cloud Machine Learning and TensorFlow 1.x to develop and evaluate prediction models using machine learning. TensorFlow is an open source, powerful, portable machine learning library developed by Google that can work with very large datasets.
This lab follows on from previous labs in this series where you created a basic prediction model using logistic regression with Spark and Pig and then used Cloud Dataflow to create training and test datasets using a pipeline that can also be used for prediction thereby eliminating the risk of training-serving skew with your prediction models. This lab is based on Chapter 9, Machine Learning Classifier Using TensorFlow, of the Data Science on the Google Cloud Platform book from O'Reilly Media, Inc.
You initially start by creating an experimental framework in Python to replicate the basic linear regression model using TensorFlow 1.x, and then expand this framework to evaluate models with more variables. This framework can then be used to compare the performance of these basic machine learning models to more complex machine learning models.
As with all of the other labs in this series, this lab provides useful experience working with data processing and data modelling techniques that will allow you to develop a data model that will predict whether flights will arrive at their destinations late given the specific details of the departure such as location, time of departure, actual departure delay, and taxi out time, etc.
The data set that is used provides historic information about internal flights in the United States retrieved from the US Bureau of Transport Statistics website. This data set can be used to demonstrate a wide range of data science concepts and techniques and is used in all of the other labs in the Data Science on Google Cloud: Machine Learning quest. The specific datasets used for this lab are the aggregated datasets developed in the previous lab in this quest, Processing Time Windowed Data with Apache Beam and Cloud Dataflow (Java).
- Develop a basic TensorFlow 1.x experimental framework in Python
- Extend the framework to create a Linear Classifier model with additional features
- Deploy the experimental training and evaluation framework to Google Cloud ML
Join Qwiklabs to read the rest of this lab...and more!
- Get temporary access to the Google Cloud Console.
- Over 200 labs from beginner to advanced levels.
- Bite-sized so you can learn at your own pace.