algorithm

How to build privacy and security into deep learning models

How to build privacy and security into deep learning models

December 31, 2019

In recent years, we've seen tremendous improvements in artificial intelligence, due to the advances of neural-based models. However, the more popular these algorithms and techniques get, the more serious the consequences of data and user privacy. Yishay Carmiel reviews these issues and explains how they impact the future of deep learning development.
Leveraging AI for social good

Leveraging AI for social good

December 30, 2019

The hardware, software, and algorithms that automatically tag our images or recommend the next book to read can also improve medical diagnosis and protect our natural resources. Jack Dashwood and Anna Bethke discuss a variety of technical projects at Intel that have enabled social good organizations and provide guidance on creating or engaging in these types of projects.
Synergize your tech stack to realize AIs full potential (sponsored by SAS)

Synergize your tech stack to realize AIs full potential (sponsored by SAS)

December 28, 2019

The whole of AI is greater than the sum of its parts, but achieving the best analytics edge often requires a mixture of technologieschaining together AI technologies to build smart end-to-end processes. Katie Taylor explores use cases within key industries to uncover how companies are succeeding with AI through a layered technology stack.
Analytics Zoo: Distributed TensorFlow in production on Apache Spark

Analytics Zoo: Distributed TensorFlow in production on Apache Spark

December 27, 2019

Yuhao Yang and Jennie Wang demonstrate how to run distributed TensorFlow on Apache Spark with the open source software package Analytics Zoo. Compared to other solutions, Analytics Zoo is built for production environments and encourages more industry users to run deep learning applications with the big data ecosystems.
Applications of mixed effects random forests

Applications of mixed effects random forests

December 27, 2019

Clustered data is all around us. The best way to attack it? Mixed effect models. Sourav Dey explains how the mixed effects random forests (MERF) model and Python package marries the world of classical mixed effect modeling with modern machine learning algorithms and shows how it can be extended to be used with other advanced modeling techniques like gradient boosting machines and deep learning.
Deploying data science for national economic statistics

Deploying data science for national economic statistics

December 25, 2019

Jeff Chen shares strategies for overcoming time series challenges at the intersection of macroeconomics and data science, drawing from machine learning research conducted at the Bureau of Economic Analysis aimed at improving its flagship product the gross domestic product.
Efficient multi-armed bandit with Thompson sampling for applications with delayed feedback

Efficient multi-armed bandit with Thompson sampling for applications with delayed feedback

December 25, 2019

Decision making often struggles with the exploration-exploitation dilemma. Multi-armed bandits (MAB) are a popular reinforcement learning solution, but increasing the number of decision criteria leads to an exponential blowup in complexity, and observational delays dont allow for optimal performance. Shradha Agrawal offers an overview of MABs and explains how to overcome the above challenges.
Federated learning

Federated learning

December 24, 2019

Imagine building a model whose training data is collected on edge devices such as cell phones or sensors. Each device collects data unlike any other, and the data cannot leave the device because of privacy concerns or unreliable network access. This challenging situation is known as federated learning. Mike Lee Williams discusses the algorithmic solutions and the product opportunities.
New directions in record linkage

New directions in record linkage

December 22, 2019

The US Census Bureau has been involved in record linkage projects for over 40 years. In that time, there's been a lot of change in computing capabilities and new techniques, and the Census Bureau is reviewing an inventory of linkage methodologies. Yves Thibaudeau describes the progress made so far in identifying specific record linkage techniques for specific applications.
NLP from scratch: Solving the cold start problem for natural language processing

NLP from scratch: Solving the cold start problem for natural language processing

December 22, 2019

How do you train a machine learning model with no training data? Michael Johnson and Norris Heintzelman share their journey implementing multiple solutions to bootstrapping training data in the NLP domain, covering topics including weak supervision, building an active learning framework, and annotation adjudication for named-entity recognition.
The magic behind your Lyft ride prices: A case study on machine learning and streaming

The magic behind your Lyft ride prices: A case study on machine learning and streaming

December 20, 2019

Rakesh Kumar and Thomas Weise explore how Lyft dynamically prices its rides with a combination of various data sources, ML models, and streaming infrastructure for low latency, reliability, and scalabilityallowing the pricing system to be more adaptable to real-world changes.
User-based real-time product recommendations leveraging deep learning using Analytics Zoo on Apache Spark and BigDL

User-based real-time product recommendations leveraging deep learning using Analytics Zoo on Apache Spark and BigDL

December 20, 2019

User-based real-time recommendation systems have become an important topic in ecommerce. Lu Wang, Nicole Kong, Guoqiong Song, and Maneesha Bhalla demonstrate how to build deep learning algorithms using Analytics Zoo with BigDL on Apache Spark and create an end-to-end system to serve real-time product recommendations.
Building a robust content recommendation platform for 60 million news readers

Building a robust content recommendation platform for 60 million news readers

December 18, 2019

Matt Chapman leads a walkthrough of the architecture and open source components that serve Tribune Publishing's content recommendation system, powered by online machine learning at scale. Find out how multiple publications, multiple recommendation algorithms, and one scalable architecture regularly achieve double the performance of the legacy solution.
Leveling Up Your CD: Unlocking Progressive Delivery on Kubernetes

Leveling Up Your CD: Unlocking Progressive Delivery on Kubernetes

December 8, 2019

Kubernetes Continuous Delivery methods have continued to evolve to more advanced strategies such as canary, A/B testing, and blue-green. Progressive delivery is the next step of CD, enabling service p …
A Toolkit for Simulating Kubernetes Scheduling at Scale

A Toolkit for Simulating Kubernetes Scheduling at Scale

December 1, 2019

As Kubernetes becomes the de facto standard for container orchestration, new scheduling algorithms and systems are made for different scenarios and workloads. Unfortunately, it is very time and cost c …
Databases on Kubernetes Using a Custom Operator: Day 1, Day 2, and Beyond

Databases on Kubernetes Using a Custom Operator: Day 1, Day 2, and Beyond

November 14, 2019

We started the journey of building a managed cloud version of the graph database Neo4j. A bit later we started developing an operator to manage multiple database clusters in k8s.Handling persistence a …
Accelerating the Journey of an AI Algorithm to Production with OpenFaaS

Accelerating the Journey of an AI Algorithm to Production with OpenFaaS

November 6, 2019

At BT it used to take us 2-3 years to get from a research idea to an AI algorithm that can be consumed by an end-user in production. We needed to reduce the risk and the investment required to bring o …
Hyperparameter Tuning Using Kubeflow

Hyperparameter Tuning Using Kubeflow

October 12, 2019

In machine learning, hyperparameter tuning refers to the process of finding the optimal constraints for training models. Choosing optimal hyperparameters can drastically improve the performance of a m …
Why is Cloud-Native Application Development Still So Hard?

Why is Cloud-Native Application Development Still So Hard?

October 5, 2019

Creating reliable, scalable, distributed cloud-native applications is still way too difficult for us mere mortal application developers. Why must we understand and solve complex distributed systems pr …
Recursive Read Deadlocks and Where to Find Them

Recursive Read Deadlocks and Where to Find Them

October 3, 2019

Lockdep (the deadlock detector in the Linux kernel) is a powerful tool to detect deadlocks, and has been used for a long time by kernel developers. However, when comes to read/write lock deadlock dete …
Network Bandwidth-Aware Kubernetes Cluster

Network Bandwidth-Aware Kubernetes Cluster

October 2, 2019

Currently Kubernetes scheduler only considers cpu and memory but with network bandwidth oversubscription widely exists in the data centers, bandwidth contention is a big problem for bandwidth sensitiv …
Protecting Sensitive Code with Encrypted Container Images on Kubernetes

Protecting Sensitive Code with Encrypted Container Images on Kubernetes

October 1, 2019

Many enterprises are driven by trade secrets in their code - whether it is a proprietary AI model, or a secret high frequency trading strategy. It is of utmost importance that critical algorithms, pro …
Anomaly Detection for Cloud Native Storage

Anomaly Detection for Cloud Native Storage

September 22, 2019

Integrating with heterogeneous storage in the Cloud Native environment has always been a challenge. How to detect problems and fix them in a timely fashion is important for mission critical workloads. …
Deep Dive: etcd

Deep Dive: etcd

September 22, 2019

etcd, a distributed key-value store, is the most critical component to Kubernetes control plane to provide strong consistency and durability of cluster metadata. etcd implements Raft consensus algorit …