Machine Learning

In our day to day life the programming concepts are changing, instead of programming a computer we can teach a computer to learn programming and work for us. ML concepts are introduced to drive unstructured data to structured manner. Once we get the data, we can easily use that data for getting useful information out of that data. So Google machine learning engine is basically a structured service that allows us to create learning models that can work on any type or size of data. We can create our models with a tensorFlow framework which powers many Google products like 'Google photo', 'Google speech' etc. This cloud machine can be deployed on any type of tensorFlow model and perform training on a cluster. It can also managing online and batch predictions. This service is managed by Google cloud dataflow for preprocessing and allow us to use data from Google cloud storage and bigquery. This machine engine allows developers to create models using cloud data lab. One can understand data, create model graphs, train their model and analyze the outcome from the model.

Example

A system is created. We show the syatem two similar looking glass filled with different liqid and ask them to identify which glass has beer and which one is filled with wine. To answer this question that machine builds a model and this model starts a process called training. Machine learning has a goal to create an actual model and use training to answer our questions. In order to create a model we need to collect data. This data can be an image, it can be text document or music tune. In beer or wine scenario, the training data can predict whether the glass has wine or beer. The data may be like the shape of the glass, the color of drink, alcohol percentage of the liquid etc.

Intro

Intro

Intro

• More

UtterlyCloud

• admin@utterlycloud.com
• India