Julien SIMON, profile picture
Uploaded byJulien SIMON
PDF, PPTX895 views

Become a Machine Learning developer with AWS services (May 2019)

This document discusses using machine learning to detect telecom fraud at Lebara, a mobile virtual network operator. It describes Lebara's current rule-based fraud detection approach and how machine learning could help automate and improve fraud detection. The document outlines Lebara's process of working with AWS experts on a workshop using Lebara call detail records to build, train, and deploy machine learning models with Amazon SageMaker to detect revenue share fraud. It also discusses the work done to prepare the data, including analyzing, cleaning, and filtering the call records for model training.

Embed presentation

Download as PDF, PPTX
Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Become a Machine Learning developer with AWS services Julien Simon Global Evangelist, AI & Machine Learning, AWS @julsimon Lars Hoogweg CTO, Lebara
Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Machine learning cycle Business Problem ML problem framing Data collection Data integration Data preparation and cleaning Data visualization and analysis Feature engineering Model training and parameter tuning Model evaluation Monitoring and debugging Model deployment Predictions Are business goals met? YESNO Dataaugmentation Feature augmentation Re-training
Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Build your dataset Business Problem ML problem framing Data collection Data integration Data preparation and cleaning Data visualization and analysis Feature engineering Model training and parameter tuning Model evaluation Monitoring and debugging Model deployment Predictions Are business goals met? YESNO Dataaugmentation Feature augmentation Re-training
Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Annotating data at scale is time-consuming and expensive
Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Amazon SageMaker Ground Truth Buildscalableandcost-effectivelabelingworkflows Easily integrate human labelers Get accurate results K E Y F E AT U R E S Automatic labeling via machine learning Ready-made and custom workflows for image bounding box, segmentation, and text Quickly label training data Private and public human workforce
Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Prepare your dataset for Machine Learning Business Problem ML problem framing Data collection Data integration Data preparation and cleaning Data visualization and analysis Feature engineering Model training and parameter tuning Model evaluation Monitoring and debugging Model deployment Predictions Are business goals met? YESNO Dataaugmentation Feature augmentation Re-training
Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Build, train and deploy models using compute services Business Problem ML problem framing Data collection Data integration Data preparation and cleaning Data visualization and analysis Feature engineering Model training and parameter tuning Model evaluation Monitoring and debugging Model deployment Predictions Are business goals met? YESNO Dataaugmentation Feature augmentation Re-training Amazon EC2 Amazon EKS Amazon ECS AWS Batch
Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Š 2018, Amazon Web Services, Inc. or Its Affiliates. All rights reserved. AWS Deep Learning AMIs Preconfigured environments on Amazon Linux or Ubuntu Conda AMI For developers who want pre- installed pip packages of DL frameworks in separate virtual environments. Base AMI For developers who want a clean slate to set up private DL engine repositories or custom builds of DL engines. AMI with source code For developers who want preinstalled DL frameworks and their source code in a shared Python environment.
Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Build, train and deploy models using SageMaker Business Problem ML problem framing Data collection Data integration Data preparation and cleaning Data visualization and analysis Feature engineering Model training and parameter tuning Model evaluation Monitoring and debugging Model deployment Predictions Are business goals met? YESNO Dataaugmentation Feature augmentation Re-training
Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Model options Training code Factorization Machines Linear Learner Principal Component Analysis K-Means Clustering XGBoost And more Built-in Algorithms (17) No ML coding required No infrastructure work required Distributed training Pipe mode Bring Your Own Container Full control, run anything! R, C++, etc. No infrastructure work required Built-in Frameworks Bring your own code: script mode Open source containers No infrastructure work required Distributed training Pipe mode
Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Using Machine Learning to detect Telco Fraud Lars Hoogweg Chief Technology Officer Lebara
Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Agenda About Lebara Telco Fraud Using ML for detecting Telco Fraud Next Steps
© 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T About Lebara • Mobile Virtual Network Operator • Active in 5 countries across Europe • Our mission: to make it easier for migrant communities to stay connected to family and friends back home
© 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T What is Telco Fraud? “the use of telecommunications products or services with the intention of illegally acquiring money from a telecommunication company or its customers”
© 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Telco Fraud Examples • SIM Boxing • A SIM box is a device containing a number of SIM cards. These SIM cards are used to terminate (international) calls bypassing international interconnect charges • One A-number calling many different B-numbers • Revenue Share Fraud • Generate traffic to high cost, revenue share service numbers • Multiple A-numbers calling the same B-number or range of B-numbers. • Higher than average call duration • Wangiri Fraud • A special case of Revenue Share Fraud • Making random calls from premium rate numbers, letting the calls ring once and then hanging up, hoping that recipients call back
© 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Fraud Detection @ Lebara • Current fraud detection approach is rule based • Fraudsters may change their patterns when they hit these rules • We cannot detect the fraud we do not know • Can we use ML to improve our fraud detection capabilities? • Automating fraud detection • Detecting new types of fraud? • How do we find out given our limited knowledge of ML?
© 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Approach • Organized a three-day offsite workshop together with AWS ML experts • Working with actual Lebara data: Call Detail Records (CDRs) • Data set labeled using existing fraud system • Three groups focusing on three different types of fraud • Focus on Revenue Share Fraud for the rest of this presentation • Training and deploying models with Amazon SageMaker
© 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Call Detail Records (CDR) • For each call, SMS, data session, top up, etc., a CDR is generated in real-time by Lebara’s Online Charging System • Lebara streams CDRs using Amazon Kinesis Firehose and stores them in Amazon S3 • So, what does a CDR look like?
Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T An Example Call Detail Record 704000001796849823|0|20190504205402|1|31616531654|31624586868|31616531654||0624586868||1| 00|316530200000|204083309537469|||20190504215245|8|0|0|20190504215254|0|68|1|304543413330 43443337||120||||1287862183|310008|31616531654|1|0|20190501|0|0|0|0||31|99|1|31|99|99999|3 1|99|2|31|99|2|310008|1000000000000000000|0|0|0||||1000000|102779300616118593|2||0|120|0|0 |192440000|0|0|102779200616117893|0||6372|0|120|179880|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0| 0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|||||||0|0|0||||||||20190501000000||||0|0|0|0|0|0|0||0|0| 0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0||||31624586868||||ocg2;1556999564;69969258;2|||||||||||1| N|N|201905041950|20190504215402|D|11000|11|102779300616118593|102779200616117893|5010000 0060964541NLD|50100000060964541NLD|1003|1|0|0|0|0|0|0|0|0|0|50100000060964541NLD|1027793 00616118593|50100000060964541NLD|0|0|0|||0|0|||0|0||||0|||0|0|||0|0||||0|||0|0|||0|0||||0||| 0|0|||0|0||||0||F|S|31616531654|704000000967825550|1003|0|179880|319903||||0|0|0|0|0|0|0||| ||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0 |0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0| 0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0||0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0 |0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0| |0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|||||||316530200000|0|0|0|1|73253560
Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T An Example Call Detail Record 704000001796849823|0|20190504205402|1|31616531654|31624586868|31616531654||0624586868||1| 00|316530200000|204083309537469|||20190504215245|8|0|0|20190504215254|0|68|1|304543413330 43443337||120||||1287862183|310008|31616531654|1|0|20190501|0|0|0|0||31|99|1|31|99|99999|3 1|99|2|31|99|2|310008|1000000000000000000|0|0|0||||1000000|102779300616118593|2||0|120|0|0 |192440000|0|0|102779200616117893|0||6372|0|120|179880|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0| 0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|||||||0|0|0||||||||20190501000000||||0|0|0|0|0|0|0||0|0| 0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0||||31624586868||||ocg2;1556999564;69969258;2|||||||||||1| N|N|201905041950|20190504215402|D|11000|11|102779300616118593|102779200616117893|5010000 0060964541NLD|50100000060964541NLD|1003|1|0|0|0|0|0|0|0|0|0|50100000060964541NLD|1027793 00616118593|50100000060964541NLD|0|0|0|||0|0|||0|0||||0|||0|0|||0|0||||0|||0|0|||0|0||||0||| 0|0|||0|0||||0||F|S|31616531654|704000000967825550|1003|0|179880|319903||||0|0|0|0|0|0|0||| ||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0 |0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0| 0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0||0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0 |0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0| |0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|||||||316530200000|0|0|0|1|73253560 Timestamp2019-05-04 21:52:54 A-number 31616531654 B-number 31624586868 Duration
© 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Data preparation • A significant amount of time was spent analyzing and preparing the data • Removing calls to non-numeric or too long B-numbers • Filtering out calls to short numbers, like IVR and CS as these are certainly not fraud and may skew the results (many A-numbers calling a few B-numbers)
© 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Feature Engineering • Creating the variables used to train the machine learning model • Features that could be used for detecting Revenue Share Fraud • Time of day / day of week • Count of different A-numbers calling a B-number range within a given time window • Ratio of A- to B-numbers • Average call duration / standard deviation
© 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Using built-in algorithms in Amazon SageMaker • Unsupervised learning for anomaly detection • Algorithm used: Random Cut Forest • Actual (previously unknown) fraud detected! • Supervised learning using our labeled dataset • A needle in a hay-stack: only 1 in every 3000 calls is considered fraudulent • Algorithm used: XGBoost • Despite the extreme unbalance, initial results are promising • Next step is tuning the model to reduce the number of false negatives Confusion Matrix Prediction Not Fraud Fraud Actual Not Fraud 114010 0 Fraud 417 187
© 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Conclusions • Using Amazon SageMaker, Lebara could get started with limited prior ML knowledge • Lebara managed to achieve promising results for detecting telco fraud within days • Besides continuing work on the fraud detection use case, we are looking at applying ML in other areas as well
Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Hands-on with Amazon SageMaker
© 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T The Amazon SageMaker API • Python SDK orchestrating all Amazon SageMaker activity • High-level objects for algorithm selection, training, deploying, model tuning, etc. • Spark SDK too (Python & Scala) • AWS SDK • For scripting and automation • CLI : ‘aws sagemaker’ • Language SDKs: boto3, etc.
Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Demo: Automatic Model Tuning with XGBoost https://gitlab.com/juliensimon/ent321/blob/master/ENT321%20- %20short%20version.ipynb
Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.S U M M I T Š 2018, Amazon Web Services, Inc. or Its Affiliates. All rights reserved. Getting started http://aws.amazon.com/free https://aws.amazon.com/sagemaker https://github.com/aws/sagemaker-python-sdk https://github.com/aws/sagemaker-spark https://github.com/awslabs/amazon-sagemaker-examples https://gitlab.com/juliensimon/ent321 https://medium.com/@julsimon https://gitlab.com/juliensimon/dlnotebooks https://gitlab.com/juliensimon/dlcontainers
Thank you! S U M M I T Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved. Julien Simon Global Evangelist, AI & Machine Learning, AWS @julsimon Lars Hoogweg CTO, Lebara
S U M M I T Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.

More Related Content

PDF
Build, train and deploy ML models with Amazon SageMaker (May 2019)
byJulien SIMON
 
PDF
Build, train and deploy Machine Learning models on Amazon SageMaker (May 2019)
byJulien SIMON
 
PPTX
Build, train and deploy ML models with SageMaker (October 2019)
byJulien SIMON
 
PPTX
Optimize your machine learning workloads on AWS (March 2019)
byJulien SIMON
 
PPTX
Build, Train and Deploy Machine Learning Models at Scale (April 2019)
byJulien SIMON
 
PDF
A pragmatic introduction to natural language processing models (October 2019)
byJulien SIMON
 
PPTX
Building Machine Learning Inference Pipelines at Scale (July 2019)
byJulien SIMON
 
PPTX
Machine Learning on AWS (December 2018)
byJulien SIMON
 
Build, train and deploy ML models with Amazon SageMaker (May 2019)
byJulien SIMON
 
Build, train and deploy Machine Learning models on Amazon SageMaker (May 2019)
byJulien SIMON
 
Build, train and deploy ML models with SageMaker (October 2019)
byJulien SIMON
 
Optimize your machine learning workloads on AWS (March 2019)
byJulien SIMON
 
Build, Train and Deploy Machine Learning Models at Scale (April 2019)
byJulien SIMON
 
A pragmatic introduction to natural language processing models (October 2019)
byJulien SIMON
 
Building Machine Learning Inference Pipelines at Scale (July 2019)
byJulien SIMON
 
Machine Learning on AWS (December 2018)
byJulien SIMON
 

What's hot

PDF
Scaling Machine Learning from zero to millions of users (May 2019)
byJulien SIMON
 
PDF
Get started with Machine Learning and Computer Vision Using AWS DeepLens (Feb...
byJulien SIMON
 
PDF
Build Machine Learning Models with Amazon SageMaker (April 2019)
byJulien SIMON
 
PDF
Starting your AI/ML project right (May 2020)
byJulien SIMON
 
PPTX
Amazon SageMaker (December 2018)
byJulien SIMON
 
PDF
Building smart applications with AWS AI services (October 2019)
byJulien SIMON
 
PPTX
Automate your Amazon SageMaker Workflows (July 2019)
byJulien SIMON
 
PPTX
The Future of AI (September 2019)
byJulien SIMON
 
PPTX
Building machine learning inference pipelines at scale (March 2019)
byJulien SIMON
 
PDF
AIM410R Deep Learning Applications with TensorFlow, featuring Mobileye (Decem...
byJulien SIMON
 
PPTX
Deep Learning on Amazon Sagemaker (July 2019)
byJulien SIMON
 
PPTX
AWS re:Invent 2018 - AIM302 - Machine Learning at the Edge
byJulien SIMON
 
PPTX
Deep Learning with TensorFlow and Apache MXNet on Amazon SageMaker (March 2019)
byJulien SIMON
 
PPTX
Optimize your Machine Learning workloads (April 2019)
byJulien SIMON
 
PPTX
AWS re:Invent 2018 - AIM401 - Deep Learning using Tensorflow
byJulien SIMON
 
PPTX
Aws autopilot
byVivek Raja P S
 
Scaling Machine Learning from zero to millions of users (May 2019)
byJulien SIMON
 
Get started with Machine Learning and Computer Vision Using AWS DeepLens (Feb...
byJulien SIMON
 
Build Machine Learning Models with Amazon SageMaker (April 2019)
byJulien SIMON
 
Starting your AI/ML project right (May 2020)
byJulien SIMON
 
Amazon SageMaker (December 2018)
byJulien SIMON
 
Building smart applications with AWS AI services (October 2019)
byJulien SIMON
 
Automate your Amazon SageMaker Workflows (July 2019)
byJulien SIMON
 
The Future of AI (September 2019)
byJulien SIMON
 
Building machine learning inference pipelines at scale (March 2019)
byJulien SIMON
 
AIM410R Deep Learning Applications with TensorFlow, featuring Mobileye (Decem...
byJulien SIMON
 
Deep Learning on Amazon Sagemaker (July 2019)
byJulien SIMON
 
AWS re:Invent 2018 - AIM302 - Machine Learning at the Edge
byJulien SIMON
 
Deep Learning with TensorFlow and Apache MXNet on Amazon SageMaker (March 2019)
byJulien SIMON
 
Optimize your Machine Learning workloads (April 2019)
byJulien SIMON
 
AWS re:Invent 2018 - AIM401 - Deep Learning using Tensorflow
byJulien SIMON
 
Aws autopilot
byVivek Raja P S
 

Similar to Become a Machine Learning developer with AWS services (May 2019)

PDF
Mcl345 re invent_sagemaker_dmbanga
byDan Romuald Mbanga
 
PDF
Unleash the Power of ML with AWS | AWS Summit Tel Aviv 2019
byAWS Summits
 
PDF
AWS Summit Singapore 2019 | Build, Train and Deploy Deep Learning Models on A...
byAWS Summits
 
PDF
Amazon의 머신러닝 솔루션: Fraud Detection & Predictive Maintenance - 남궁영환 (AWS 데이터 사이...
byAmazon Web Services Korea
 
PPTX
WhereML a Serverless ML Powered Location Guessing Twitter Bot
byRandall Hunt
 
PPTX
Where ml ai_heavy
byRandall Hunt
 
PDF
Amazon reInvent 2020 Recap: AI and Machine Learning
byChris Fregly
 
PDF
AWS reinvent 2019 recap - Riyadh - AI And ML - Ahmed Raafat
byAWS Riyadh User Group
 
PDF
Artificial intelligence in actions: delivering a new experience to Formula 1 ...
byGoDataDriven
 
PDF
Become a Machine Learning developer with AWS (Avril 2019)
byJulien SIMON
 
PDF
Real-World Smart Applications with Amazon Machine Learning - AWS Machine Lear...
byAWS Germany
 
PDF
Building a Production-ready Predictive App for Customer Service - Alex Ingerm...
byPAPIs.io
 
PDF
Innovation Track AWS Cloud Experience Argentina - Democratizing Artificial In...
byAmazon Web Services LATAM
 
PPTX
Applying ML on your Data in Motion with AWS and Confluent | Joseph Morais, Co...
byHostedbyConfluent
 
PPTX
Drive Digital Transformation Using Machine Learning
byAWS Summits
 
PDF
[AWS Techshift] Innovation and AI/ML Sagemaker Build-in 머신러닝 모델 활용 및 Marketpl...
byAmazon Web Services Korea
 
PPTX
AWS Toronto Summit 2019 - AIM302 - Build, train, and deploy ML models with Am...
byJonathan Dion
 
PDF
SKL208 - Turbocharge your Business with AI and Machine Learning - Tel Aviv Su...
byBoaz Ziniman
 
PDF
Machine Learning for everyone
byJulien SIMON
 
PDF
MLops workshop AWS
byGili Nachum
 
Mcl345 re invent_sagemaker_dmbanga
byDan Romuald Mbanga
 
Unleash the Power of ML with AWS | AWS Summit Tel Aviv 2019
byAWS Summits
 
AWS Summit Singapore 2019 | Build, Train and Deploy Deep Learning Models on A...
byAWS Summits
 
Amazon의 머신러닝 솔루션: Fraud Detection & Predictive Maintenance - 남궁영환 (AWS 데이터 사이...
byAmazon Web Services Korea
 
WhereML a Serverless ML Powered Location Guessing Twitter Bot
byRandall Hunt
 
Where ml ai_heavy
byRandall Hunt
 
Amazon reInvent 2020 Recap: AI and Machine Learning
byChris Fregly
 
AWS reinvent 2019 recap - Riyadh - AI And ML - Ahmed Raafat
byAWS Riyadh User Group
 
Artificial intelligence in actions: delivering a new experience to Formula 1 ...
byGoDataDriven
 
Become a Machine Learning developer with AWS (Avril 2019)
byJulien SIMON
 
Real-World Smart Applications with Amazon Machine Learning - AWS Machine Lear...
byAWS Germany
 
Building a Production-ready Predictive App for Customer Service - Alex Ingerm...
byPAPIs.io
 
Innovation Track AWS Cloud Experience Argentina - Democratizing Artificial In...
byAmazon Web Services LATAM
 
Applying ML on your Data in Motion with AWS and Confluent | Joseph Morais, Co...
byHostedbyConfluent
 
Drive Digital Transformation Using Machine Learning
byAWS Summits
 
[AWS Techshift] Innovation and AI/ML Sagemaker Build-in 머신러닝 모델 활용 및 Marketpl...
byAmazon Web Services Korea
 
AWS Toronto Summit 2019 - AIM302 - Build, train, and deploy ML models with Am...
byJonathan Dion
 
SKL208 - Turbocharge your Business with AI and Machine Learning - Tel Aviv Su...
byBoaz Ziniman
 
Machine Learning for everyone
byJulien SIMON
 
MLops workshop AWS
byGili Nachum
 

More from Julien SIMON

PDF
Implementing high-quality and cost-effiient AI applications with small langua...
byJulien SIMON
 
PDF
Optimiser vos workloads AI/ML sur Amazon EC2 et AWS Graviton
byJulien SIMON
 
PDF
Trying to figure out MCP by actually building an app from scratch with open s...
byJulien SIMON
 
PDF
Arcee AI - building and working with small language models (06/25)
byJulien SIMON
 
PDF
deep_dive_multihead_latent_attention.pdf
byJulien SIMON
 
PDF
Deep Dive: Model Distillation with DistillKit
byJulien SIMON
 
PDF
Deep Dive: Parameter-Efficient Model Adaptation with LoRA and Spectrum
byJulien SIMON
 
PDF
Building High-Quality Domain-Specific Models with Mergekit
byJulien SIMON
 
PDF
Tailoring Small Language Models for Enterprise Use Cases
byJulien SIMON
 
PDF
Tailoring Small Language Models for Enterprise Use Cases
byJulien SIMON
 
PDF
Julien Simon - Deep Dive: Compiling Deep Learning Models
byJulien SIMON
 
PDF
Tailoring Small Language Models for Enterprise Use Cases
byJulien SIMON
 
PDF
Julien Simon - Deep Dive - Optimizing LLM Inference
byJulien SIMON
 
PDF
Julien Simon - Deep Dive - Accelerating Models with Better Attention Layers
byJulien SIMON
 
PDF
Julien Simon - Deep Dive - Quantizing LLMs
byJulien SIMON
 
PDF
Julien Simon - Deep Dive - Model Merging
byJulien SIMON
 
PDF
An introduction to computer vision with Hugging Face
byJulien SIMON
 
PDF
Reinventing Deep Learning
 with Hugging Face Transformers
byJulien SIMON
 
PDF
Building NLP applications with Transformers
byJulien SIMON
 
PPTX
Building Machine Learning Models Automatically (June 2020)
byJulien SIMON
 
Implementing high-quality and cost-effiient AI applications with small langua...
byJulien SIMON
 
Optimiser vos workloads AI/ML sur Amazon EC2 et AWS Graviton
byJulien SIMON
 
Trying to figure out MCP by actually building an app from scratch with open s...
byJulien SIMON
 
Arcee AI - building and working with small language models (06/25)
byJulien SIMON
 
deep_dive_multihead_latent_attention.pdf
byJulien SIMON
 
Deep Dive: Model Distillation with DistillKit
byJulien SIMON
 
Deep Dive: Parameter-Efficient Model Adaptation with LoRA and Spectrum
byJulien SIMON
 
Building High-Quality Domain-Specific Models with Mergekit
byJulien SIMON
 
Tailoring Small Language Models for Enterprise Use Cases
byJulien SIMON
 
Tailoring Small Language Models for Enterprise Use Cases
byJulien SIMON
 
Julien Simon - Deep Dive: Compiling Deep Learning Models
byJulien SIMON
 
Tailoring Small Language Models for Enterprise Use Cases
byJulien SIMON
 
Julien Simon - Deep Dive - Optimizing LLM Inference
byJulien SIMON
 
Julien Simon - Deep Dive - Accelerating Models with Better Attention Layers
byJulien SIMON
 
Julien Simon - Deep Dive - Quantizing LLMs
byJulien SIMON
 
Julien Simon - Deep Dive - Model Merging
byJulien SIMON
 
An introduction to computer vision with Hugging Face
byJulien SIMON
 
Reinventing Deep Learning
 with Hugging Face Transformers
byJulien SIMON
 
Building NLP applications with Transformers
byJulien SIMON
 
Building Machine Learning Models Automatically (June 2020)
byJulien SIMON
 

Recently uploaded

PDF
Our Digital Tribe_ Cultivating Connection and Growth in Our Slack Community 🌿...
bysanjeetmishra30
 
PDF
Access Control 2025: From Security Silo to Software-Defined Ecosystem
byMemoori
 
PPTX
Software Analysis &Design ethiopia chap-2.pptx
byAbbas484771
 
PPTX
Conversational Agents – Building Intelligent Assistants [Virtual Hands-on Wor...
byUiPathCommunity
 
PDF
December Patch Tuesday
byIvanti
 
PDF
The year in review - MarvelClient in 2025
bypanagenda
 
PDF
Day 3 - Data and Application Security - 2nd Sight Lab Cloud Security Class
by2nd Sight Lab
 
PPTX
AI in Cybersecurity: Digital Defense by Yasir Naveed Riaz
byYasir Naveed Riaz
 
PDF
Vibe Coding vs. Spec-Driven Development [Free Meetup]
byHaim Michael
 
PPTX
Cybersecurity Best Practices - Step by Step guidelines
byYasir Naveed Riaz
 
PPTX
cybercrime in Information security .pptx
byismaeelsahib032
 
PPTX
Building Cyber Resilience for 2026: Best Practices for a Secure, AI-Driven Bu...
byYasir Naveed Riaz
 
PDF
API-First Architecture in Financial Systems
bycyy111112
 
PDF
Cybersecurity: Safeguarding Digital Assets
byYasir Naveed Riaz
 
PPTX
From Backup to Resilience: How MSPs Are Preparing for 2026
byMSP360
 
PDF
Six Shifts For 2026 (And The Next Six Years)
byDavid Armano
 
PDF
Internet_of_Things_IoT_for_Next_Generation_Smart_Systems_Utilizing.pdf
byRoshanSyed12
 
PPTX
DYNAMICALLY.pptx good for the teachers or students to do seminars and for tea...
bympoorvika031
 
PDF
TrustArc Webinar - Looking Ahead: The 2026 Privacy Landscape
byTrustArc
 
PPT
software-security-intro in information security.ppt
byismaeelsahib032
 
Our Digital Tribe_ Cultivating Connection and Growth in Our Slack Community 🌿...
bysanjeetmishra30
 
Access Control 2025: From Security Silo to Software-Defined Ecosystem
byMemoori
 
Software Analysis &Design ethiopia chap-2.pptx
byAbbas484771
 
Conversational Agents – Building Intelligent Assistants [Virtual Hands-on Wor...
byUiPathCommunity
 
December Patch Tuesday
byIvanti
 
The year in review - MarvelClient in 2025
bypanagenda
 
Day 3 - Data and Application Security - 2nd Sight Lab Cloud Security Class
by2nd Sight Lab
 
AI in Cybersecurity: Digital Defense by Yasir Naveed Riaz
byYasir Naveed Riaz
 
Vibe Coding vs. Spec-Driven Development [Free Meetup]
byHaim Michael
 
Cybersecurity Best Practices - Step by Step guidelines
byYasir Naveed Riaz
 
cybercrime in Information security .pptx
byismaeelsahib032
 
Building Cyber Resilience for 2026: Best Practices for a Secure, AI-Driven Bu...
byYasir Naveed Riaz
 
API-First Architecture in Financial Systems
bycyy111112
 
Cybersecurity: Safeguarding Digital Assets
byYasir Naveed Riaz
 
From Backup to Resilience: How MSPs Are Preparing for 2026
byMSP360
 
Six Shifts For 2026 (And The Next Six Years)
byDavid Armano
 
Internet_of_Things_IoT_for_Next_Generation_Smart_Systems_Utilizing.pdf
byRoshanSyed12
 
DYNAMICALLY.pptx good for the teachers or students to do seminars and for tea...
bympoorvika031
 
TrustArc Webinar - Looking Ahead: The 2026 Privacy Landscape
byTrustArc
 
software-security-intro in information security.ppt
byismaeelsahib032
 

Become a Machine Learning developer with AWS services (May 2019)

  • 1.
    Š 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Become a Machine Learning developer with AWS services Julien Simon Global Evangelist, AI & Machine Learning, AWS @julsimon Lars Hoogweg CTO, Lebara
  • 2.
    Š 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Machine learning cycle Business Problem ML problem framing Data collection Data integration Data preparation and cleaning Data visualization and analysis Feature engineering Model training and parameter tuning Model evaluation Monitoring and debugging Model deployment Predictions Are business goals met? YESNO Dataaugmentation Feature augmentation Re-training
  • 3.
    Š 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Build your dataset Business Problem ML problem framing Data collection Data integration Data preparation and cleaning Data visualization and analysis Feature engineering Model training and parameter tuning Model evaluation Monitoring and debugging Model deployment Predictions Are business goals met? YESNO Dataaugmentation Feature augmentation Re-training
  • 4.
    Š 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Annotating data at scale is time-consuming and expensive
  • 5.
    Š 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Amazon SageMaker Ground Truth Buildscalableandcost-effectivelabelingworkflows Easily integrate human labelers Get accurate results K E Y F E AT U R E S Automatic labeling via machine learning Ready-made and custom workflows for image bounding box, segmentation, and text Quickly label training data Private and public human workforce
  • 6.
    Š 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Prepare your dataset for Machine Learning Business Problem ML problem framing Data collection Data integration Data preparation and cleaning Data visualization and analysis Feature engineering Model training and parameter tuning Model evaluation Monitoring and debugging Model deployment Predictions Are business goals met? YESNO Dataaugmentation Feature augmentation Re-training
  • 7.
    Š 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Build, train and deploy models using compute services Business Problem ML problem framing Data collection Data integration Data preparation and cleaning Data visualization and analysis Feature engineering Model training and parameter tuning Model evaluation Monitoring and debugging Model deployment Predictions Are business goals met? YESNO Dataaugmentation Feature augmentation Re-training Amazon EC2 Amazon EKS Amazon ECS AWS Batch
  • 8.
    Š 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Š 2018, Amazon Web Services, Inc. or Its Affiliates. All rights reserved. AWS Deep Learning AMIs Preconfigured environments on Amazon Linux or Ubuntu Conda AMI For developers who want pre- installed pip packages of DL frameworks in separate virtual environments. Base AMI For developers who want a clean slate to set up private DL engine repositories or custom builds of DL engines. AMI with source code For developers who want preinstalled DL frameworks and their source code in a shared Python environment.
  • 9.
    Š 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Build, train and deploy models using SageMaker Business Problem ML problem framing Data collection Data integration Data preparation and cleaning Data visualization and analysis Feature engineering Model training and parameter tuning Model evaluation Monitoring and debugging Model deployment Predictions Are business goals met? YESNO Dataaugmentation Feature augmentation Re-training
  • 10.
    Š 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Model options Training code Factorization Machines Linear Learner Principal Component Analysis K-Means Clustering XGBoost And more Built-in Algorithms (17) No ML coding required No infrastructure work required Distributed training Pipe mode Bring Your Own Container Full control, run anything! R, C++, etc. No infrastructure work required Built-in Frameworks Bring your own code: script mode Open source containers No infrastructure work required Distributed training Pipe mode
  • 11.
    Š 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Using Machine Learning to detect Telco Fraud Lars Hoogweg Chief Technology Officer Lebara
  • 12.
    Š 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Agenda About Lebara Telco Fraud Using ML for detecting Telco Fraud Next Steps
  • 13.
    © 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T About Lebara • Mobile Virtual Network Operator • Active in 5 countries across Europe • Our mission: to make it easier for migrant communities to stay connected to family and friends back home
  • 14.
    © 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T What is Telco Fraud? “the use of telecommunications products or services with the intention of illegally acquiring money from a telecommunication company or its customers”
  • 15.
    © 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Telco Fraud Examples • SIM Boxing • A SIM box is a device containing a number of SIM cards. These SIM cards are used to terminate (international) calls bypassing international interconnect charges • One A-number calling many different B-numbers • Revenue Share Fraud • Generate traffic to high cost, revenue share service numbers • Multiple A-numbers calling the same B-number or range of B-numbers. • Higher than average call duration • Wangiri Fraud • A special case of Revenue Share Fraud • Making random calls from premium rate numbers, letting the calls ring once and then hanging up, hoping that recipients call back
  • 16.
    © 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Fraud Detection @ Lebara • Current fraud detection approach is rule based • Fraudsters may change their patterns when they hit these rules • We cannot detect the fraud we do not know • Can we use ML to improve our fraud detection capabilities? • Automating fraud detection • Detecting new types of fraud? • How do we find out given our limited knowledge of ML?
  • 17.
    © 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Approach • Organized a three-day offsite workshop together with AWS ML experts • Working with actual Lebara data: Call Detail Records (CDRs) • Data set labeled using existing fraud system • Three groups focusing on three different types of fraud • Focus on Revenue Share Fraud for the rest of this presentation • Training and deploying models with Amazon SageMaker
  • 18.
    © 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Call Detail Records (CDR) • For each call, SMS, data session, top up, etc., a CDR is generated in real-time by Lebara’s Online Charging System • Lebara streams CDRs using Amazon Kinesis Firehose and stores them in Amazon S3 • So, what does a CDR look like?
  • 19.
    Š 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T An Example Call Detail Record 704000001796849823|0|20190504205402|1|31616531654|31624586868|31616531654||0624586868||1| 00|316530200000|204083309537469|||20190504215245|8|0|0|20190504215254|0|68|1|304543413330 43443337||120||||1287862183|310008|31616531654|1|0|20190501|0|0|0|0||31|99|1|31|99|99999|3 1|99|2|31|99|2|310008|1000000000000000000|0|0|0||||1000000|102779300616118593|2||0|120|0|0 |192440000|0|0|102779200616117893|0||6372|0|120|179880|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0| 0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|||||||0|0|0||||||||20190501000000||||0|0|0|0|0|0|0||0|0| 0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0||||31624586868||||ocg2;1556999564;69969258;2|||||||||||1| N|N|201905041950|20190504215402|D|11000|11|102779300616118593|102779200616117893|5010000 0060964541NLD|50100000060964541NLD|1003|1|0|0|0|0|0|0|0|0|0|50100000060964541NLD|1027793 00616118593|50100000060964541NLD|0|0|0|||0|0|||0|0||||0|||0|0|||0|0||||0|||0|0|||0|0||||0||| 0|0|||0|0||||0||F|S|31616531654|704000000967825550|1003|0|179880|319903||||0|0|0|0|0|0|0||| ||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0 |0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0| 0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0||0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0 |0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0| |0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|||||||316530200000|0|0|0|1|73253560
  • 20.
    Š 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T An Example Call Detail Record 704000001796849823|0|20190504205402|1|31616531654|31624586868|31616531654||0624586868||1| 00|316530200000|204083309537469|||20190504215245|8|0|0|20190504215254|0|68|1|304543413330 43443337||120||||1287862183|310008|31616531654|1|0|20190501|0|0|0|0||31|99|1|31|99|99999|3 1|99|2|31|99|2|310008|1000000000000000000|0|0|0||||1000000|102779300616118593|2||0|120|0|0 |192440000|0|0|102779200616117893|0||6372|0|120|179880|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0| 0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|||||||0|0|0||||||||20190501000000||||0|0|0|0|0|0|0||0|0| 0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0|0||||31624586868||||ocg2;1556999564;69969258;2|||||||||||1| N|N|201905041950|20190504215402|D|11000|11|102779300616118593|102779200616117893|5010000 0060964541NLD|50100000060964541NLD|1003|1|0|0|0|0|0|0|0|0|0|50100000060964541NLD|1027793 00616118593|50100000060964541NLD|0|0|0|||0|0|||0|0||||0|||0|0|||0|0||||0|||0|0|||0|0||||0||| 0|0|||0|0||||0||F|S|31616531654|704000000967825550|1003|0|179880|319903||||0|0|0|0|0|0|0||| ||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0 |0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0| 0|0|0|0|0|||||0|0|0|0|0|0|0|||||0|0|0|0|0|0|0||0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0 |0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0| |0|0|0||0|0|0|0|0||0|0|0||0|0|0|0|0||0|||||||316530200000|0|0|0|1|73253560 Timestamp2019-05-04 21:52:54 A-number 31616531654 B-number 31624586868 Duration
  • 21.
    © 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Data preparation • A significant amount of time was spent analyzing and preparing the data • Removing calls to non-numeric or too long B-numbers • Filtering out calls to short numbers, like IVR and CS as these are certainly not fraud and may skew the results (many A-numbers calling a few B-numbers)
  • 22.
    © 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Feature Engineering • Creating the variables used to train the machine learning model • Features that could be used for detecting Revenue Share Fraud • Time of day / day of week • Count of different A-numbers calling a B-number range within a given time window • Ratio of A- to B-numbers • Average call duration / standard deviation
  • 23.
    © 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Using built-in algorithms in Amazon SageMaker • Unsupervised learning for anomaly detection • Algorithm used: Random Cut Forest • Actual (previously unknown) fraud detected! • Supervised learning using our labeled dataset • A needle in a hay-stack: only 1 in every 3000 calls is considered fraudulent • Algorithm used: XGBoost • Despite the extreme unbalance, initial results are promising • Next step is tuning the model to reduce the number of false negatives Confusion Matrix Prediction Not Fraud Fraud Actual Not Fraud 114010 0 Fraud 417 187
  • 24.
    © 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Conclusions • Using Amazon SageMaker, Lebara could get started with limited prior ML knowledge • Lebara managed to achieve promising results for detecting telco fraud within days • Besides continuing work on the fraud detection use case, we are looking at applying ML in other areas as well
  • 25.
    Š 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Hands-on with Amazon SageMaker
  • 26.
    © 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T The Amazon SageMaker API • Python SDK orchestrating all Amazon SageMaker activity • High-level objects for algorithm selection, training, deploying, model tuning, etc. • Spark SDK too (Python & Scala) • AWS SDK • For scripting and automation • CLI : ‘aws sagemaker’ • Language SDKs: boto3, etc.
  • 27.
    Š 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Demo: Automatic Model Tuning with XGBoost https://gitlab.com/juliensimon/ent321/blob/master/ENT321%20- %20short%20version.ipynb
  • 28.
    Š 2019, AmazonWeb Services, Inc. or its affiliates. All rights reserved.S U M M I T Š 2018, Amazon Web Services, Inc. or Its Affiliates. All rights reserved. Getting started http://aws.amazon.com/free https://aws.amazon.com/sagemaker https://github.com/aws/sagemaker-python-sdk https://github.com/aws/sagemaker-spark https://github.com/awslabs/amazon-sagemaker-examples https://gitlab.com/juliensimon/ent321 https://medium.com/@julsimon https://gitlab.com/juliensimon/dlnotebooks https://gitlab.com/juliensimon/dlcontainers
  • 29.
    Thank you! S UM M I T Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved. Julien Simon Global Evangelist, AI & Machine Learning, AWS @julsimon Lars Hoogweg CTO, Lebara
  • 30.
    S U MM I T Š 2019, Amazon Web Services, Inc. or its affiliates. All rights reserved.