This document discusses t-digest, which provides a compact way to represent a distribution of values. T-digest uses adaptive bins that are smaller near the edges, allowing it to accurately track quantiles even with a limited number of bins. It works by taking data samples, sorting them, and grouping them into bins while respecting a maximum size. The bins can then be merged across samples or time periods. T-digest is useful for applications that need to track distributions over many variables or time periods with limited space.
This talk shows practical methods for find changes in a variety of kinds of data as well as giving real-world examples from finance, telecom, systems monitoring and natural language processing.
Tensor Abuse - how to reuse machine learning frameworksTed Dunning
Tensors are a very useful tool for mathematical programming. Moreover, the optimization frameworks that are part of most machine learning frameworks have some very cool uses outside of the normal machine learning kinds of tasks.
The logistics of machine learning typically take waaay more effort than the machine learning itself. Moreover, machine learning systems aren't like normal software projects so continuous integration takes on new meaning.
Real-time Puppies and Ponies - Evolving Indicator Recommendations in Real-timeTed Dunning
This talk describes how indicator-based recommendations can be evolved in real time. Normally, indicator-based recommendations use a large off-line computation to understand the general structure of items to be recommended and then make recommendations in real-time to users based on a comparison of their recent history versus the large-scale product of the off-line computation.
In this talk, I show how the same components of the off-line computation that guarantee linear scalability in a batch setting also give strict real-time bounds on the cost of a practical real-time implementation of the indicator computation.
This talk focuses on how larger data sets are not only enabling advanced techniques, but also increasing the number of problems within reach of relatively simple techniques, that is "cheap learning".
This was one of the talks that I gave at the Strata San Jose conference. I migrated my topic a bit, but here is the original abstract:
Application developers and architects today are interested in making their applications as real-time as possible. To make an application respond to events as they happen, developers need a reliable way to move data as it is generated across different systems, one event at a time. In other words, these applications need messaging.
Messaging solutions have existed for a long time. However, when compared to legacy systems, newer solutions like Apache Kafka offer higher performance, more scalability, and better integration with the Hadoop ecosystem. Kafka and similar systems are based on drastically different assumptions than legacy systems and have vastly different architectures. But do these benefits outweigh any tradeoffs in functionality? Ted Dunning dives into the architectural details and tradeoffs of both legacy and new messaging solutions to find the ideal messaging system for Hadoop.
Topics include:
* Queues versus logs
* Security issues like authentication, authorization, and encryption
* Scalability and performance
* Handling applications that span multiple data centers
* Multitenancy considerations
* APIs, integration points, and more
Many statistics are impossible to compute precisely on streaming data. There are some very clever algorithms, however, which allow us to compute very good approximations of these values efficiently in terms of CPU and memory.
This talk shows practical methods for find changes in a variety of kinds of data as well as giving real-world examples from finance, telecom, systems monitoring and natural language processing.
Tensor Abuse - how to reuse machine learning frameworksTed Dunning
Tensors are a very useful tool for mathematical programming. Moreover, the optimization frameworks that are part of most machine learning frameworks have some very cool uses outside of the normal machine learning kinds of tasks.
The logistics of machine learning typically take waaay more effort than the machine learning itself. Moreover, machine learning systems aren't like normal software projects so continuous integration takes on new meaning.
Real-time Puppies and Ponies - Evolving Indicator Recommendations in Real-timeTed Dunning
This talk describes how indicator-based recommendations can be evolved in real time. Normally, indicator-based recommendations use a large off-line computation to understand the general structure of items to be recommended and then make recommendations in real-time to users based on a comparison of their recent history versus the large-scale product of the off-line computation.
In this talk, I show how the same components of the off-line computation that guarantee linear scalability in a batch setting also give strict real-time bounds on the cost of a practical real-time implementation of the indicator computation.
This talk focuses on how larger data sets are not only enabling advanced techniques, but also increasing the number of problems within reach of relatively simple techniques, that is "cheap learning".
This was one of the talks that I gave at the Strata San Jose conference. I migrated my topic a bit, but here is the original abstract:
Application developers and architects today are interested in making their applications as real-time as possible. To make an application respond to events as they happen, developers need a reliable way to move data as it is generated across different systems, one event at a time. In other words, these applications need messaging.
Messaging solutions have existed for a long time. However, when compared to legacy systems, newer solutions like Apache Kafka offer higher performance, more scalability, and better integration with the Hadoop ecosystem. Kafka and similar systems are based on drastically different assumptions than legacy systems and have vastly different architectures. But do these benefits outweigh any tradeoffs in functionality? Ted Dunning dives into the architectural details and tradeoffs of both legacy and new messaging solutions to find the ideal messaging system for Hadoop.
Topics include:
* Queues versus logs
* Security issues like authentication, authorization, and encryption
* Scalability and performance
* Handling applications that span multiple data centers
* Multitenancy considerations
* APIs, integration points, and more
Many statistics are impossible to compute precisely on streaming data. There are some very clever algorithms, however, which allow us to compute very good approximations of these values efficiently in terms of CPU and memory.
These are the slides from my talk at FAR Con in Minneapolis recently. The topics are the implications of buried treasure hoards on data security, horror stories and new, simpler and provably secure methods for public data disclosure.
Anomaly Detection - New York Machine LearningTed Dunning
Anomaly detection is the art of finding what you don't know how to ask for. In this talk, I walk through the why and how of building probabilistic models for a variety of problems including continuous signals and web traffic. This talk blends theory and practice in a highly approachable way.
Cognitive computing with big data, high tech and low tech approachesTed Dunning
I explain some very approachable methods for analyzing big data via a detour through clipper ships and the 19th century open source scene.
Note that I mixed up the route of the Flying Cloud record in this talk. The Flying Cloud's record was actually from New York to San Francisco and was even more impressive than what I said. The usual time had been about 180 days. With Maury's charts, the time was reduced to about 135 days. The Flying Cloud's time was 89 days.
Thanks to Chen Kung for noticing my error.
This is the position talk that I gave at CIKM. Included are 4 algorithms that I feel don't get much academic attention, but which are very important industrially. It isn't necessarily true that these algorithms *should* get academic attention, but I do feel that it is true that they are quite important pragmatically speaking.
This talk describes the general architecture common to anomaly detections systems that are based on probabilistic models. By examining several realistic use cases, I illustrate the common themes and practical implementation methods.
Machine Learning for Chickens, Autonomous Driving and a 3-year-old Who Won’t ...MapR Technologies
Big data technologies are being applied to a wide variety of use cases. We will review tangible examples of machine learning, discuss an autonomous driving project and illustrate the role of MapR in next generation initiatives. More: http://info.mapr.com/WB_Machine-Learning-for-Chickens_Global_DG_17.11.02_RegistrationPage.html
Apache Mahout is changing radically. Here is a report on what is coming, notably including an R like domain specific language that can use multiple computational engines such as Spark.
Building multi-modal recommendation engines using search enginesTed Dunning
This is my strata NY talk about how to build recommendation engines using common items. In particular, I show how multi-modal recommendations can be built using the same framework.
Using Mahout and a Search Engine for RecommendationTed Dunning
I presented this talk at the Open World Forum in Paris in 2013. The ideas here are that you can do basic recommendations and extended forms of recommendation such as intelligent search or cross recommendation or multi-modal recommendation using Mahout's cooccurrence analysis together with a search engine.
Recent work in recommendations allows some really amazing simplicity of implementation while extending the inputs handled to multiple kinds of interactions against items different from the ones being recommended.
ML Workshop 2: Machine Learning Model Comparison & EvaluationMapR Technologies
How Rendezvous Architecture Improves Evaluation in the Real World
In this addition of our machine learning logistics webinar series we build on the ideas of the key requirements for effective management of machine learning logistics presented in the Overview webinar and in Part I Workshop. Here we focus on model-to-model comparison & evaluation, use of decoy models and more. Listen here: http://info.mapr.com/machine-learning-workshop2.html?_ga=2.35695522.324200644.1511891424-416597139.1465233415
Spark and MapR Streams: A Motivating ExampleIan Downard
Businesses are discovering the untapped potential of large datasets and data streams through the use of technologies for big data processing and storage. By leveraging these assets they’re creating a new generation of applications that derive value from data they used to throw away. In this presentation Ian Downard shows how to build operational environments for these types of applications with the MapR Converged Data Platform and he describes examples of a next-generation applications that use Java APIs for MapR Streams, Apache Spark, Apache Hive, and MapR-DB. He shows how these technologies can be used to join and transform unbounded datasets to find signals and derive new data streams for a financial scenario involving real-time algorithmic trading and historical analysis using SQL. He also discusses how MapR enables you to run real-time data applications with the speed, reliability, and security you need for a production environment.
These are the slides from my talk at FAR Con in Minneapolis recently. The topics are the implications of buried treasure hoards on data security, horror stories and new, simpler and provably secure methods for public data disclosure.
Anomaly Detection - New York Machine LearningTed Dunning
Anomaly detection is the art of finding what you don't know how to ask for. In this talk, I walk through the why and how of building probabilistic models for a variety of problems including continuous signals and web traffic. This talk blends theory and practice in a highly approachable way.
Cognitive computing with big data, high tech and low tech approachesTed Dunning
I explain some very approachable methods for analyzing big data via a detour through clipper ships and the 19th century open source scene.
Note that I mixed up the route of the Flying Cloud record in this talk. The Flying Cloud's record was actually from New York to San Francisco and was even more impressive than what I said. The usual time had been about 180 days. With Maury's charts, the time was reduced to about 135 days. The Flying Cloud's time was 89 days.
Thanks to Chen Kung for noticing my error.
This is the position talk that I gave at CIKM. Included are 4 algorithms that I feel don't get much academic attention, but which are very important industrially. It isn't necessarily true that these algorithms *should* get academic attention, but I do feel that it is true that they are quite important pragmatically speaking.
This talk describes the general architecture common to anomaly detections systems that are based on probabilistic models. By examining several realistic use cases, I illustrate the common themes and practical implementation methods.
Machine Learning for Chickens, Autonomous Driving and a 3-year-old Who Won’t ...MapR Technologies
Big data technologies are being applied to a wide variety of use cases. We will review tangible examples of machine learning, discuss an autonomous driving project and illustrate the role of MapR in next generation initiatives. More: http://info.mapr.com/WB_Machine-Learning-for-Chickens_Global_DG_17.11.02_RegistrationPage.html
Apache Mahout is changing radically. Here is a report on what is coming, notably including an R like domain specific language that can use multiple computational engines such as Spark.
Building multi-modal recommendation engines using search enginesTed Dunning
This is my strata NY talk about how to build recommendation engines using common items. In particular, I show how multi-modal recommendations can be built using the same framework.
Using Mahout and a Search Engine for RecommendationTed Dunning
I presented this talk at the Open World Forum in Paris in 2013. The ideas here are that you can do basic recommendations and extended forms of recommendation such as intelligent search or cross recommendation or multi-modal recommendation using Mahout's cooccurrence analysis together with a search engine.
Recent work in recommendations allows some really amazing simplicity of implementation while extending the inputs handled to multiple kinds of interactions against items different from the ones being recommended.
ML Workshop 2: Machine Learning Model Comparison & EvaluationMapR Technologies
How Rendezvous Architecture Improves Evaluation in the Real World
In this addition of our machine learning logistics webinar series we build on the ideas of the key requirements for effective management of machine learning logistics presented in the Overview webinar and in Part I Workshop. Here we focus on model-to-model comparison & evaluation, use of decoy models and more. Listen here: http://info.mapr.com/machine-learning-workshop2.html?_ga=2.35695522.324200644.1511891424-416597139.1465233415
Spark and MapR Streams: A Motivating ExampleIan Downard
Businesses are discovering the untapped potential of large datasets and data streams through the use of technologies for big data processing and storage. By leveraging these assets they’re creating a new generation of applications that derive value from data they used to throw away. In this presentation Ian Downard shows how to build operational environments for these types of applications with the MapR Converged Data Platform and he describes examples of a next-generation applications that use Java APIs for MapR Streams, Apache Spark, Apache Hive, and MapR-DB. He shows how these technologies can be used to join and transform unbounded datasets to find signals and derive new data streams for a financial scenario involving real-time algorithmic trading and historical analysis using SQL. He also discusses how MapR enables you to run real-time data applications with the speed, reliability, and security you need for a production environment.
Changes in how business is done combined with multiple technology drivers make geo-distributed data increasingly important for enterprises. These changes are causing serious disruption across a wide range of industries, including healthcare, manufacturing, automotive, telecommunications, and entertainment. Technical challenges arise with these disruptions, but the good news is there are now innovative solutions to address these problems. http://info.mapr.com/WB_Geo-distributed-Big-Data-and-Analytics_Global_DG_17.05.16_RegistrationPage.html
The Hive Think Tank: Rendezvous Architecture Makes Machine Learning Logistics...The Hive
Think Tank Event 10/23/2017, hosted by The Hive and presented by Ted Dunning, Chief Application Architect of MapR Technologies and Ellen Friedman of MapR Technologies.
ML Workshop 1: A New Architecture for Machine Learning LogisticsMapR Technologies
Having heard the high-level rationale for the rendezvous architecture in the introduction to this series, we will now dig in deeper to talk about how and why the pieces fit together. In terms of components, we will cover why streams work, why they need to be persistent, performant and pervasive in a microservices design and how they provide isolation between components. From there, we will talk about some of the details of the implementation of a rendezvous architecture including discussion of when the architecture is applicable, key components of message content and how failures and upgrades are handled. We will touch on the monitoring requirements for a rendezvous system but will save the analysis of the recorded data for later. Listen to the webinar on demand: https://mapr.com/resources/webinars/machine-learning-workshop-1/
Predictive Maintenance Using Recurrent Neural NetworksJustin Brandenburg
My presentation from AnacondaCON 2018 where I discussed using Recurrent Neural Networks, Python, Tensorflow and the MapR Platform to develop deploy a predictive maintenance model for an IoT device in the manufacturing industry.
Machine Learning Success: The Key to Easier Model ManagementMapR Technologies
Join Ellen Friedman, co-author (with Ted Dunning) of a new short O’Reilly book Machine Learning Logistics: Model Management in the Real World, to look at what you can do to have effective model management, including the role of stream-first architecture, containers, a microservices approach and a DataOps style of work. Ellen will provide a basic explanation of a new architecture that not only leverages stream transport but also makes use of canary models and decoy models for accurate model evaluation and for efficient and rapid deployment of new models in production.
Ted Dunning, Chief Application Architect, MapR at MLconf SFMLconf
Abstract: Near real-time Updates for Cooccurrence-based Recommenders
Most recommendation algorithms are inherently batch oriented and require all relevant history to be processed. In some contexts such as music, this does not cause significant problems because waiting a day or three before recommendations are available for new items doesn’t significantly change their impact. In other contexts, the value of items drops precipitously with time so that recommending day-old items has little value to users.
In this talk, I will describe how a large-scale multi-modal cooccurrence recommender can be extended to include near real-time updates. In addition, I will show how these real-time updates are compatible with delivery of recommendations via search engines.
Ellen Friedman and I spoke at the ACM meetup about how stream-first architecture can have a big impact and how the logistics of machine learning is a great example of that impact.
This is my half of the presentation.
State of the Art Robot Predictive Maintenance with Real-time Sensor DataMathieu Dumoulin
Our Strata Beijing 2017 presentation slides where we show how to use data from a movement sensor, in real-time, to do anomaly detection at scale using standard enterprise big data software.
How the Internet of Things is Turning the Internet Upside DownTed Dunning
This is a wide-ranging talk that goes into how the internet is architected, how that architecture is changing as a result of internet of things, how the internet of things worked in the 19th century big data, open-source community and how to build time-series databases to make this all possible.
Really.
We introduce the idea that metadata, including project information, data labels, data characteristics and indications of valuable use, can be propagated through a data processing lineage graph. Further, finding examples of significant cooccurrence of propagated and original metadata gives us the basis of an interesting kind of search engine gives interesting recommendations of data given a problem statement even in a near cold-start situation.
The folk wisdom has always been that when running stateful applications inside containers, the only viable choice is to externalize the state so that the containers themselves are stateless or nearly so. Keeping large amounts of state inside containers is possible, but it’s considered a problem because stateful containers generally can’t preserve that state across restarts.
In practice, this complicates the management of large-scale Kubernetes-based infrastructure because these high-performance storage systems require separate management. In terms of overall system management, it would be ideal if we could run a software-defined storage system directly in containers managed by Kubernetes, but that has been hampered by lack of direct device access and difficult questions about what happens to the state on container restarts.
Ted Dunning describes recent developments that make it possible for Kubernetes to manage both compute and storage tiers in the same cluster. Container restarts can be handled gracefully without loss of data or a requirement to rebuild storage structures and access to storage from compute containers is extremely fast. In some environments, it’s even possible to implement elastic storage frameworks that can fold data onto just a few containers during quiescent periods or explode it in just a few seconds across a large number of machines when higher speed access is required.
The benefits of systems like this extend beyond management simplicity, because applications can be more Agile precisely because the storage layer is more stable and can be uniformly accessed from any container host. Even better, it makes it a snap to configure and deploy a full-scale compute and storage infrastructure.
Apache Kylin - OLAP Cubes for SQL on HadoopTed Dunning
Apache Kylin (incubating) is a new project to bring OLAP cubes to Hadoop. I walk through the project and describe how it works and how users see the project.
These are the slides that we used to ignite the conversation with the audience at Hadoop Summit EU. Come over to the Mahout dev list to be part of the ongoing conversation.
Levelwise PageRank with Loop-Based Dead End Handling Strategy : SHORT REPORT ...Subhajit Sahu
Abstract — Levelwise PageRank is an alternative method of PageRank computation which decomposes the input graph into a directed acyclic block-graph of strongly connected components, and processes them in topological order, one level at a time. This enables calculation for ranks in a distributed fashion without per-iteration communication, unlike the standard method where all vertices are processed in each iteration. It however comes with a precondition of the absence of dead ends in the input graph. Here, the native non-distributed performance of Levelwise PageRank was compared against Monolithic PageRank on a CPU as well as a GPU. To ensure a fair comparison, Monolithic PageRank was also performed on a graph where vertices were split by components. Results indicate that Levelwise PageRank is about as fast as Monolithic PageRank on the CPU, but quite a bit slower on the GPU. Slowdown on the GPU is likely caused by a large submission of small workloads, and expected to be non-issue when the computation is performed on massive graphs.
Adjusting primitives for graph : SHORT REPORT / NOTESSubhajit Sahu
Graph algorithms, like PageRank Compressed Sparse Row (CSR) is an adjacency-list based graph representation that is
Multiply with different modes (map)
1. Performance of sequential execution based vs OpenMP based vector multiply.
2. Comparing various launch configs for CUDA based vector multiply.
Sum with different storage types (reduce)
1. Performance of vector element sum using float vs bfloat16 as the storage type.
Sum with different modes (reduce)
1. Performance of sequential execution based vs OpenMP based vector element sum.
2. Performance of memcpy vs in-place based CUDA based vector element sum.
3. Comparing various launch configs for CUDA based vector element sum (memcpy).
4. Comparing various launch configs for CUDA based vector element sum (in-place).
Sum with in-place strategies of CUDA mode (reduce)
1. Comparing various launch configs for CUDA based vector element sum (in-place).
Unleashing the Power of Data_ Choosing a Trusted Analytics Platform.pdfEnterprise Wired
In this guide, we'll explore the key considerations and features to look for when choosing a Trusted analytics platform that meets your organization's needs and delivers actionable intelligence you can trust.
Chatty Kathy - UNC Bootcamp Final Project Presentation - Final Version - 5.23...John Andrews
SlideShare Description for "Chatty Kathy - UNC Bootcamp Final Project Presentation"
Title: Chatty Kathy: Enhancing Physical Activity Among Older Adults
Description:
Discover how Chatty Kathy, an innovative project developed at the UNC Bootcamp, aims to tackle the challenge of low physical activity among older adults. Our AI-driven solution uses peer interaction to boost and sustain exercise levels, significantly improving health outcomes. This presentation covers our problem statement, the rationale behind Chatty Kathy, synthetic data and persona creation, model performance metrics, a visual demonstration of the project, and potential future developments. Join us for an insightful Q&A session to explore the potential of this groundbreaking project.
Project Team: Jay Requarth, Jana Avery, John Andrews, Dr. Dick Davis II, Nee Buntoum, Nam Yeongjin & Mat Nicholas
STATATHON: Unleashing the Power of Statistics in a 48-Hour Knowledge Extravag...sameer shah
"Join us for STATATHON, a dynamic 2-day event dedicated to exploring statistical knowledge and its real-world applications. From theory to practice, participants engage in intensive learning sessions, workshops, and challenges, fostering a deeper understanding of statistical methodologies and their significance in various fields."
ViewShift: Hassle-free Dynamic Policy Enforcement for Every Data LakeWalaa Eldin Moustafa
Dynamic policy enforcement is becoming an increasingly important topic in today’s world where data privacy and compliance is a top priority for companies, individuals, and regulators alike. In these slides, we discuss how LinkedIn implements a powerful dynamic policy enforcement engine, called ViewShift, and integrates it within its data lake. We show the query engine architecture and how catalog implementations can automatically route table resolutions to compliance-enforcing SQL views. Such views have a set of very interesting properties: (1) They are auto-generated from declarative data annotations. (2) They respect user-level consent and preferences (3) They are context-aware, encoding a different set of transformations for different use cases (4) They are portable; while the SQL logic is only implemented in one SQL dialect, it is accessible in all engines.
#SQL #Views #Privacy #Compliance #DataLake
The Building Blocks of QuestDB, a Time Series Databasejavier ramirez
Talk Delivered at Valencia Codes Meetup 2024-06.
Traditionally, databases have treated timestamps just as another data type. However, when performing real-time analytics, timestamps should be first class citizens and we need rich time semantics to get the most out of our data. We also need to deal with ever growing datasets while keeping performant, which is as fun as it sounds.
It is no wonder time-series databases are now more popular than ever before. Join me in this session to learn about the internal architecture and building blocks of QuestDB, an open source time-series database designed for speed. We will also review a history of some of the changes we have gone over the past two years to deal with late and unordered data, non-blocking writes, read-replicas, or faster batch ingestion.
Learn SQL from basic queries to Advance queriesmanishkhaire30
Dive into the world of data analysis with our comprehensive guide on mastering SQL! This presentation offers a practical approach to learning SQL, focusing on real-world applications and hands-on practice. Whether you're a beginner or looking to sharpen your skills, this guide provides the tools you need to extract, analyze, and interpret data effectively.
Key Highlights:
Foundations of SQL: Understand the basics of SQL, including data retrieval, filtering, and aggregation.
Advanced Queries: Learn to craft complex queries to uncover deep insights from your data.
Data Trends and Patterns: Discover how to identify and interpret trends and patterns in your datasets.
Practical Examples: Follow step-by-step examples to apply SQL techniques in real-world scenarios.
Actionable Insights: Gain the skills to derive actionable insights that drive informed decision-making.
Join us on this journey to enhance your data analysis capabilities and unlock the full potential of SQL. Perfect for data enthusiasts, analysts, and anyone eager to harness the power of data!
#DataAnalysis #SQL #LearningSQL #DataInsights #DataScience #Analytics