Implementation of Google Prediction Api with Example

What is Prediction API?

In Prediction API,there is not an option for None of the Above.

Prediction is to know the future before it happens. Prediction in technical terms, analyzing the set of data to find out whether it is going to happen or not happen. For Example, Let’s consider the Temperature. If I am analyzing the temperature or the same day from last 10 years and I try to predict the temperature in very simple terms it is all I say prediction.

And what is the Machine Learning Terms? Training your machine to predict the data.

Examples

These all things you can do with the prediction:

1. Attachment in Email
2. Google Webline
3. Document and Email Classification
4. Customer Sentiment analysis
5. Message Routing Decisions

When I am trying to send an email using my Gmail account and if have written “Please find the attachment” and in case you forgot the attachment, Gmail will ask you that you mention that you have an attachment but you didn’t so you might forget the attachment.So, that is how Gmail is predicting, as it found the word “attachment”,and realize that in the past people are using the word means they are attaching the document.So, it is checking and asking that are you attaching the document or not.

Though sometimes, this feature becomes annoying. For Example, Let’s take the example of Google Webline Services. Google Webline Service automatically loads when your internet connection is slow and what happens,In that case, JavaScript won’t work and that is pretty much annoying because you can not switch because the connection is very slow.

It is a machine learning as service on the cloud. As we have the platform as service on the cloud(PAAS), Infrastructure as a Service(IAAS) on the cloud, Software as a service(SAAS) on the cloud. Which changing the markets at this stage are Machine Learning As A Service(mLaaS) on cloud and another thing is mobile backend as a service on the cloud (MBAAS). These Packages are going to change the future Laptops are going to smaller and smaller and soon they will be vanish In the meantime mobile devices are getting powerful and bigger.In the past it was totally reverse laptops are getting faster but now all of us wants portability.That is the future and mLaaS and mBaas.

Other things Prediction output by the based upon you enter like It can predict the language you have written and also your sentiments.

Go to the www.google.com/killer-robots.txt. You will get the following text on that page.

Here T is for The Terminator

Larry Page and Sergey Brin has left messages for T-1000 and T-800.The killer-robots.txt file prevents T-800 and T-1000 from killing Larry Page and Sergey Brin. In case,If Google never creates the system called Skynet will everyone would accept.

Next feature is machine learning and the availability on demand RestFul WebService. Do you want things on demand,right?

What Google Prediction API can do?

1. Spam Detection
2. Diagnosis
3. Suspicious Activity Identification
4. Chorn Analysis

It is “Machine Learning as Service(MLAS) on Cloud.Predicts an output value based upon entered input like language , sentiments, tags etc.Available an on-demand HTTP RESTful Web Services.

What do I do now?

Do Prediction Every day.

Modules in Prediction API

There are two modules in Prediction API

1. Free Module
2. Paid Module

Starting right from churning a small file of few kbs to a big data set of 1.5 GBs. You can fetch them all in the form of CSV. You can find it on https://cloud.google.com. If you don’t have an account, Create the account and use it. You will get 300 $ credits for 3 months.

How does Google Prediction API Work?

Three simple things to follow for Google Prediction API.

1. Upload
2. Train
3. Predict

Upload your data whatever you have, train your system and predict the results and that is what you want.And life’s Good! Don’t waste it behind too much code!!! 😀

If you want to analyse it, Go to the link and just upload random amount of data and check what type of prediction it do.All the things are open source, You can go and get all SDKs from the URL and just check out Prediction API.

Let’s Understand the Prediction API in depth:

1. Upload:

Have your data in form of CSV file and now remember when to say CSV(It doesn’t mean Excel)

I hope you know uploading is the pretty simple process. Use Google Cloud Storage and then fill up the details ,select, and app engine location and you are good to go.Prediction API does not Provide GUI

2. Train:

Create a new model using the training data to have an accurate prediction.

This Prediction API finds the relevant feature from the dataset while the training process.

POST prediction/U1.1/training/

3. Predict:

There is some type of Prediction google provides, let’s see types of Prediction.

Types of Prediction:

1. Regression
2. Classification

In a nutshell, It’s just upload,train and predict.

Usage of Prediction:

1. Email Filtering
2. Transaction Risk
3. Career Counseling
4. Message Routing
5. Diagnosis

Example

Here is the Hello World Example of Google Prediction API.The video tutorial will help you to get the live view of Google Prediction API. Let’s understand it:

Conclusion:

Google has a set of API for developers and Prediction API is one of the best API Google Provides because it make life a much simpler and also free Module in it. Now, Start to use the Google Prediction API for your work and share your experience with us.

Final Prediction: So, What do you predict today?