O'Reilly on Data

DECEMBER 12, 2019

In addition to newer innovations, the practice borrows from model risk management, traditional model diagnostics, and software testing. Because ML models can react in very surprising ways to data they’ve never seen before, it’s safest to test all of your ML models with sensitivity analysis. [9] If so, have fun debugging! [1]

Machine Learning 191

Machine Learning Modeling Testing Risk Management 191

IBM Big Data Hub

OCTOBER 16, 2023

Containers have increased in popularity and adoption ever since the release of Docker in 2013, an open-source platform for building, deploying and managing containerized applications. Containerization helps DevOps teams avoid the complications that arise when moving software from testing to production.

Data Processing 83

Data Processing Dashboards Metrics Software 83

Smart Data Collective

SEPTEMBER 4, 2019

Maternal mortality rates in the United States jumped over 25% between 2000 and 2013. Big data is also being fed into databases from social media, blood tests, doctor visits, email, forums and a variety of other mediums. Understaffed hospitals and medical errors are causing most of the deaths.

Big Data 76

Big Data Risk Testing Technology 76

DMBS2

JUNE 25, 2018

Let me start by saying: Most fundamentally, the Carpenter decision was based on and implicitly reaffirms the Katz test.* The Katz test basically says that that an individual’s right to privacy is whatever society regards as a reasonable expectation of privacy at that time. **The Disagreement with the Katz test.

Testing 61

Testing IT 61

CIO Business Intelligence

MAY 18, 2023

In The Phoenix Project: A Novel About IT, DevOps, and Helping Your Business Win (IT Revolution Press, 2013 ) , Bill — an IT manager — takes over a critical project that’s over budget and behind schedule. We need to really understand the drivers that influence customer and employee trust, as this is increasingly a litmus test,” says Johnson.

IT 130

IT Software Recreation/Entertainment Technology 130

Sisense

APRIL 2, 2019

With DevOps aiming for greater efficiencies between software development and IT teams, it was clear that automated processes needed to be put in place to build, test, and release software faster and more reliably. 25th, 2013, New Relic News and Products. Automation & Speed. Diagram: Need Frequent and High Quality Releases?

Software 80

Software Testing Enterprise IT 80

datapine

APRIL 14, 2022

In 2017 the company wanted to take its shopping experience one step further by creating an augmented reality app that allowed users to test a product without having to leave their homes. In 2013, they took a slight risk and introduced a veggie smoothie to their previously fruit-only smoothie menu. Behind the scenes. Behind the scenes.

Big Data 244

Big Data Recreation/Entertainment Analytics Data-driven 244

Howard Dresner

FEBRUARY 11, 2014

Collaboration BI At one of my weekly #BIWisdom tweetchats this month, collaboration, social media and text analytics turned up in a discussion about 2013 BI predictions that didn’t pan out. Vendors need to automate and decrease that effort.” • “I tested a social analytics tool; I was less than impressed.

Analytics 45

Analytics Business Intelligence Data Processing Marketing 45

Domino Data Lab

DECEMBER 1, 2019

Step 4: Generate the test, train and noisy MNIST data sets. Generate the train and test sets (x_train, _), (x_test, _) = mnist.load_data() x_train = x_train.astype('float32') / 255. A Survey on Concept Drift Adaptation” ACM Computing Survey Volume 1 , Article 1 (January 2013). x_test = x_test.astype('float32') / 255.

Modeling 77

Modeling Machine Learning Deep Learning Testing 77

BizAcuity

DECEMBER 31, 2019

With the limited time F1 drivers have to test these days, rookies have to learn things very quickly – and he has done it very well. Killingsworth had 10 rebounds but they were outrebounded 48 – and none was more significant than the one on Brown’s missed free throw in the final seconds. I grew up on St.

Reporting 52

Reporting Testing Modeling IT 52

CIO Business Intelligence

NOVEMBER 7, 2022

When Target was launching in Canada in 2013, though, they assumed they would avoid this problem: there would be no data to convert, just new information to input into their SAP system. An investigation found that only about 30% of the data in the system was actually correct. Surviving an ERP rollout So what have we learned?

Manufacturing 128

Manufacturing Cost-Benefit Sales Consulting 128

bridgei2i

JULY 28, 2020

It is my immense pleasure to introduce you all to our guest today Ria Persad, she’s named as international woman of the year by Renewable Energy World in power engineering in 2013 and the lifetime achievement leader by Platts Global Energy awards in 2014. We need people who can test.

Enterprise 98

Enterprise Digital Transformation Insurance B2B 98

DataKitchen

DECEMBER 30, 2020

In June of 2020, Database Trends & Applications featured DataKitchen’s end-to-end DataOps platform for its ability to coordinate data teams, tools, and environments in the entire data analytics organization with features such as meta-orchestration , automated testing and monitoring , and continuous deployment : DataKitchen [link].

Testing 241

Testing Big Data Statistics Manufacturing 241

The Data Visualisation Catalogue

MARCH 18, 2019

Here, 3D Treemaps were tested for visualising the classificatory distribution of patent collections in the International Patent Classification (IPC) system. Schumann Rendering of Complex 3D Treemaps (GRAPP 2013). 3D Treemaps have also been experimented with in the visualisation of intellectual property. Luboschnik, H.

Visualization 47

Visualization Interactive Software Experimentation 47

Insight

MARCH 12, 2020

In 2013, Robert Galbraith?—?an I tested several different flavors of BERT for use as synopsis classifiers before settling on the DistilBERT model from Hugging Face. On my test set, this approach resulted in~75–95% accuracy and ~.65 Test Case: Dune Let’s see an example of genre tag prediction in action.

Modeling 59

Modeling Metadata Publishing Sales 59

Cloudera

AUGUST 6, 2018

For example, Crisis Text Line , which provides online support to people in crisis, received a total of 8 m illion text messages in the first two years of its existence between 2013 and 2015. Fox Foundation is testing a watch-type wearable device in Australia to continuously monitor the symptoms of patients with Parkinson’s disease.

Machine Learning 49

Machine Learning Internet of Things Cost-Benefit Data-driven 49

Howard Dresner

AUGUST 27, 2013

Before the perfect storm, our tweetchat tribe (comprised of customers, vendors and consultants/analysts) were of the opinion that the growing “app” mentality for “cool stuff” among consumers and the easy-to-consume info in mobile apps could end up increasing trust and thus lead to less testing and faster releases.

Business Intelligence 60

Business Intelligence Finance Risk Data Processing 60

AWS Big Data

NOVEMBER 14, 2023

Wallapop’s initial data architecture platform Wallapop is a Spanish ecommerce marketplace company focused on second-hand items, founded in 2013. Since its creation in 2013, it has reached more than 40 million downloads and more than 700 million products have been listed. The marketplace can be accessed via mobile app or website.

Data Warehouse 59

Data Warehouse Analytics Testing Cost-Benefit 59

CIO Business Intelligence

APRIL 4, 2024

Delete old data I lived in Hong Kong 2008-2013 and one of my most pleasurable weekends was a trip to see an incredible band at the MGM hotel in Macau. Firstly, it tests your ability to destroy data. Twelve years later, in September 2023, some of my details were compromised in the MGM Resorts hack in the USA.

Cost-Benefit 98

Cost-Benefit Data Governance Testing Modeling 98

CIO Business Intelligence

FEBRUARY 10, 2023

But for two years, we were testing limits within the public cloud.” Randich, who came to FINRA.org in 2013 after stints as co-CIO of Citigroup and former CIO of Nasdaq, is no stranger to the public cloud. “We spent about a year and a half going through several bottlenecks, taking them out one at a time with Amazon engineers.

Unstructured Data 105

Unstructured Data Data Lake Machine Learning Enterprise 105

Sisense

MAY 26, 2020

While training a model for NLP, words not present in the training data commonly appear in the test data. Because of this, predictions made using test data may not be correct. Using the semantic meaning of words it already knows as a base, the model can understand the meanings of words it doesn’t know that appear in test data.

Unstructured Data 98

Unstructured Data Big Data Testing Machine Learning 98

IBM Big Data Hub

FEBRUARY 26, 2024

With this expanded observability, incidents can be prevented in the design phase or identified in the implementation and testing phase to reduce maintenance costs and achieve higher productivity. The Basel Committee released BCBS 239 as far back as 2013.

Cost-Benefit 106

Cost-Benefit Metadata Data Governance Reporting 106

The Unofficial Google Data Science Blog

APRIL 23, 2024

Multiparameter experiments, however, generate richer data than standard A/B tests, and automated t-tests alone are insufficient to analyze them well. We use PrePost in most of our A/B tests, so we have pre-experiment metric measurements readily available that we can use as covariates in our models. Springer Netherlands, 2013. [16]

Experimentation 52

Experimentation Optimization Uncertainty Metrics 52

AWS Big Data

MARCH 13, 2024

Ray cluster for ingestion and creating vector embeddings In our testing, we found that the GPUs make the biggest impact to performance when creating the embeddings. He entered the big data space in 2013 and continues to explore that area. Therefore, we decided to use a Ray cluster to convert our raw text and create the embeddings.

Data Processing 107

Data Processing Dashboards Machine Learning Management 107

IBM Big Data Hub

FEBRUARY 22, 2024

Reliable “The Department’s AI capabilities will have explicit, well-defined uses, and the safety, security, and effectiveness of such capabilities will be subject to testing and assurance within those defined uses across their entire life cycles.” This is misguided. But it is well worth the effort.

Metadata 107

Metadata Measurement Risk Modeling 107

Insight

MAY 20, 2020

Once you’ve successfully installed Spark on to your cluster, be sure to check out the open-source “Quick Start” guide for a basic walk through of the interactive Spark shell (spark-shell)— which is a great way to learn about Spark commands and test some of its functionality interactively.

Data Warehouse 106

Data Warehouse Optimization Interactive Cost-Benefit 106

Domino Data Lab

APRIL 21, 2021

The dataset contains transactions made by European credit card holders in September 2013, and has been anonymized – Features V1, V2, …, V28 are results from applying PCA on the raw data. This is to prevent any information leakage into our test set. 2f%% of the test set." 2f%% of the test set."

Statistics 69

Statistics Machine Learning Modeling Metrics 69

AWS Big Data

FEBRUARY 1, 2024

Data collected after 2013 is stored in WARC format and includes corresponding metadata (WAT) and text extraction data (WET). Using an EMR on EC2 cluster can help you carry out tests before submitting jobs to the production environment. In this section, we cover common ways to interact, filter, and process the Common Crawl dataset.

Metadata 92

Metadata Modeling Data Processing Unstructured Data 92

AWS Big Data

NOVEMBER 29, 2023

In 2013, Amazon Web Services revolutionized the data warehousing industry by launching Amazon Redshift , the first fully-managed, petabyte-scale, enterprise-grade cloud data warehouse. Amazon Redshift made it simple and cost-effective to efficiently analyze large volumes of data using existing business intelligence tools.

Data Warehouse 97

Data Warehouse Data Lake Analytics Machine Learning 97

The Unofficial Google Data Science Blog

NOVEMBER 4, 2015

One way to check $f_theta$ is to gather test data and check whether the model fits the relationship between training and test data. This tests the model’s ability to distinguish what is common for each item between the two data sets (the underlying $theta$) and what is different (the draw from $f_theta$).

KDD 40

KDD Testing Machine Learning Modeling 40

Occam's Razor

JANUARY 31, 2017

In our case, every table, every slide that comes from a piece of data, has to pass the so what test. Great question… This last piece of data, how much video advertising's influence has changed between 2013 and 2015 in the individual making their final choice, is killer. The so what was $80 million in missed revenue.

Visualization 148

Visualization IT Strategy Key Performance Indicator 148

Occam's Razor

DECEMBER 3, 2012

It is not just what you do to attract traffic (what most people think of as marketing and advertising), but also what types of experiences you create (something people rarely think is marketing) and how good you are at delivering for where you should be in 2013 rather than 2009 (only the rarest of marketers think with this lens on).

Marketing 147

Marketing Advertising Strategy Optimization 147

BizAcuity

MAY 10, 2022

2013: Google launches Google Compute Engine (IaaS), its own version of EC2. AWS rolls out SageMaker, designed to build, train, test and deploy machine learning (ML) models. After a year of battling rough waters, Heroku was sailing with the wind. The platform was acquired by Salesforce in 2010 for over USD $200 million.

Data-driven 52

Data-driven IoT Unstructured Data Data Lake 52

Smart Data Collective

MARCH 22, 2019

By 2020, the robotics technology market is expected to reach almost $83 bn, according to the Global Robotics Technology Market 2013-2020. Test the suggestions in a single department and scale the implemented technologies factory-wide, if successful. As a comparison, the 2015 expenditure of robotics was only $71 bn. Conclusion.

Manufacturing 46

Manufacturing Cost-Benefit Enterprise Technology 46

Occam's Razor

FEBRUARY 3, 2014

According to Nielsen, YouTube reaches more US adults ages 18-34 than any cable network as of mid-2013. As of March 2013, one billion, (B!), One more thing to ponder… One hundred hours of video is uploaded into YouTube every single minute, as of May 2013. YouTube Marketing and Analytics Framework for Success.

Marketing 149

Marketing Analytics Advertising Measurement 149

Edwin Chen

AUGUST 14, 2014

How can we understand causal lifts in the absence of an A/B test? Assume that it's impossible to run an A/B test now that the feature has been launched.) For simplicity, let's assume that the badge was given to all users who received at least 5000 upvotes in 2013. Propensity Modeling. a negative effect).

Modeling 75

Modeling Measurement Sales Testing 75

Occam's Razor

DECEMBER 9, 2013

One that reflects the customer expectations of 2013. Or Ford (it is amazing that in 2013, for such an expensive product, it looks so… 2005). Go buy a simple A/B testing tool ( Visual Website Optimizer , Optimizely ), go crazy optimizing every little thing to take money from the people who want to give it to you!

Marketing 165

Marketing Analytics Key Performance Indicator Optimization 165