README.md

Training the Spam Filter model with Amazon SageMaker

Overview

Amazon SageMaker is a fully-managed service that enables developers and data scientists to quickly and easily build, train, and deploy machine learning models at any scale. Amazon SageMaker removes all the barriers that typically slow down developers who want to use machine learning.

Machine learning often feels a lot harder than it should be to most developers because the process to build and train models, and then deploy them into production is too complicated and too slow. First, you need to collect and prepare your training data to discover which elements of your data set are important. Then, you need to select which algorithm and framework you’ll use. After deciding on your approach, you need to teach the model how to make predictions by training, which requires a lot of compute. Then, you need to tune the model so it delivers the best possible predictions, which is often a tedious and manual effort. After you’ve developed a fully trained model, you need to integrate the model with your application and deploy this application on infrastructure that will scale. All of this takes a lot of specialized expertise, access to large amounts of compute and storage, and a lot of time to experiment and optimize every part of the process. In the end, it's not a surprise that the whole thing feels out of reach for most developers.

Amazon SageMaker removes the complexity that holds back developer success with each of these steps. Amazon SageMaker includes modules that can be used together or independently to build, train, and deploy your machine learning models.

In this section, we will walk you through creating and training a spam filter machine learning model with Amazon SageMaker.

The Dataset

In order to train the model, we will use the UCI SMS Spam Collection dataset. It contains a public set of SMS labeled messages that have been collected for mobile phone spam research.

Number of examples	5574
Ham	4827
Spam	747

If you want to know more about the origin of the words spam & ham please visit: https://en.wiktionary.org/wiki/ham_e-mail

The collection is composed by just one text file, where each line has the correct class followed by the raw message. Some examples are bellow:

Class	SMS Text
ham	What you doing?how are you?
ham	Ok lar... Joking wif u oni...
ham	dun say so early hor... U c already then say...
ham	MY NO. IN LUTON 0125698789 RING ME IF UR AROUND! H*
ham	Siva is in hostel aha:-.
ham	Cos i was out shopping wif darren jus now n i called him 2 ask wat present he wan lor. Then he started guessing who i was wif n he finally guessed darren lor.
spam	FreeMsg: Txt: CALL to No: 86888 & claim your reward of 3 hours talk time to use from your phone now! ubscribe6GBP/ mnth inc 3hrs 16 stop?txtStop
spam	Sunshine Quiz! Win a super Sony DVD recorder if you canname the capital of Australia? Text MQUIZ to 82277. B
spam	URGENT! Your Mobile No 07808726822 was awarded a L2,000 Bonus Caller Prize on 02/09/03! This is our 2nd attempt to contact YOU! Call 0871-872-9758 BOX95QU

The Task

We are going to build and train a binary classification model that will be able to predict whether a SMS message contains spam or not.

Building and Training the model

Create an Amazon S3 bucket

In this section, we will create an Amazon S3 bucket as staging area. Amazon SageMaker uses Amazon S3 as the main storage for both data and model artifacts; you can actually use other sources when loading data into the Jupyter notebook instances, but this is outside of the scope of this lab.

1. Sign into the AWS Management Console at https://console.aws.amazon.com/

2. In the upper-right corner of the AWS Management Console, confirm you are in the desired AWS region. For this workshop we will use the US East (N. Virginia) [us-east-1]

3. Open the Amazon S3 console at https://console.aws.amazon.com/s3 or choose the Amazon S3 service in the menu.

4. In the Amazon S3 console, click the Create Bucket button.

5. For the Bucket Name, type smlambda-workshop-[your-initials] in the text box and click Next (take note of the bucket name, it will be needed later for loading data in the notebook instance). Press Next to move to the next screen. Note: if the bucket name is already taken, feel free to add an extra suffix.

   

6. Enable versioning of the objects in the bucket as shown in the screen below. This will be required to use AWS CodePipeline in the next sections of the workshop.

   

   Press Next and then Next again leaving the settings as they are in the following screen.

7. Finally, click Create Bucket in the Review page.

Create a managed Jupyter Notebook instance

In this section we will use an Amazon SageMaker managed Jupyter notebook instance to prepare and process data and write the code to train the model. An Amazon SageMaker notebook instance is a fully managed ML compute instance running the Jupyter Notebook application. Amazon SageMaker manages creating the instance and related resources.

1. In the AWS Management Console, click on Services, type “SageMaker” and press enter.

   

2. You’ll be placed in the Amazon SageMaker dashboard. Click on Create notebook instance.

   

3. In the Create notebook instance screen

   

   1. Give the Notebook Instance a name like notebook-instance-[your-initials]

   2. Choose ml.t2.medium as Notebook instance type

   3. Choose Create a new role in the IAM role dropdown list. Notebook instances require permissions to call other services including Amazon SageMaker and Amazon S3 APIs. Choose Specific S3 buckets in the Create an IAM role window and input the name of the bucket that you have created in the previous section. Then click on Create role.

      If you already had a role with the proper grants to access the newly created bucket, creating a new role is not mandatory and you can just choose to use the existing role.

   4. Keep No VPC selected in the VPC dropdown list

   5. Keep No configuration selected in the Lifecycle configuration dropdown list

   6. Keep No Custom Encryption selected in the Encryption key dropdown list

   7. Finally, click on Create notebook instance

4. You will be redirected to the Notebook instances screen

   Wait until the notebook instance is status is In Service and then click on the Open button to be redirected to Jupyter

   

   

Download notebook and training code to the notebook instance

For the purpose of this workshop the code required to build and train the Machine Learning model is pre-implemented and available for download from GitHub.

As a consequence, in this section we will clone the GitHub repository into the Amazon SageMaker notebook instance and access the Jupyter notebook to run training.

1. Click on New > Terminal in the right-hand side of the Jupyter interface

   

   This will open a terminal window in the Jupyter interface

   

2. Execute the following commands in the terminal

   cd SageMaker
   git clone https://github.com/giuseppeporcelli/smlambdaworkshop

3. When the clone operation completes, close the terminal window and return to the Jupyter landing page. The folder smlambdaworkshop will appear automatically (if not, you can hit the Refresh button)

   

4. Browse to the folder smlambdaworkshop > training and open the file sms_spam_classifier_mxnet.ipynb

Execute notebook code and run training

We are ready to execute the steps for building and training our Machine Learning model!

1. Once the notebook file is opened, you will be asked to set a kernel. Choose conda_mxnet_p36 and click on Set Kernel. Kernels in Jupyter are responsible of code execution, providing replyies, etc. and generally allow the execution of a single programming language.

   

2. We will use the Jupyter Notebook as the interface for buiding our model and call Amazon SageMaker for training. Jupyter allows code execution by selecting the code cell and pressing on the Run button in the toolbar (alternatively, you can use the SHIFT + Enter keyboard shortcut.

   

3. Next steps are documented in the Jupyter notebook itself. From an high-level perspective, the notebook code will:

   • Download the dataset to the notebook instance
   • Execute some pre-processing to vectorize SMS messages
   • Split the dataset 80%/20% (training/validation)
   • Build a neural-network using Gluon, a deep-learning interface for Apache MXNet
   • Execute training on Amazon SageMaker

   The notebook code does not need to be modified, except for the name of the Amazon S3 bucket that you have created and used in the previous sections.

   

4. Execute all the cells in sequence (note that while the code in a cell is running the interface shows [*] to the upper-left of the cell) up to the execution of the training, which will start a Training Job in Amazon SageMaker.

   

   You will see several outputs in the notebook, showing the progress of the training process.

5. You can also monitor the status of the training job from the Amazon SageMaker console

   

6. When the training job completes successfully, you can move to the next section!

   Please note that the notebook code includes also other steps which allow to automatically deploy the trained ML model to the managed hosting service provided by Amazon SageMaker, which will create a real-time endpoint that can be invoked to get inferences. Those steps are not mandatory for this workshop and are implemented for educational purposes.