AI & Machine Learning News. 28, September 2020
The Artificial Intelligence and Machine Learning News clippings for Quants are provided algorithmically with CloudQuant’s NLP engine which seeks out articles relevant to our community and ranks them by our proprietary interest score. After all, shouldn’t you expect to see the news generated using AI?
Microsoft AI Bringing Old Photos back to life
Bringing Old Photos Back to Life, CVPR2020 (Oral)
Old Photo Restoration via Deep Latent Space Translation, PAMI Under Review
Ziyu Wan, Bo Zhang, Dongdong Chen, Pan Zhang, Dong Chen, Jing Liao, Fang Wen
2020-09-17 Read the Full Story
CloudQuant Thoughts : Very impressive Microsoft!
‘OpenAI should be renamed ClosedAI’: Reaction to Microsoft’s exclusive license of OpenAI’s GPT-3
Microsoft this week gained an exclusive license to OpenAI’s GPT-3, the state-of-the-art language model garnering attention across the tech industry. Other companies will still be able to access the model through an Azure-hosted API, but only Microsoft will have access to GPT-3’s code and underlying advances. The deal follows Microsoft’s $1 billion investment last year in San Francisco-based OpenAI, which consists of the OpenAI Inc nonprofit founded four years ago and the for-profit OpenAI LP.
The implications of giving a tech giant such as Microsoft an exclusive license to GPT-3 raises questions and potential concerns. MIT Technology Review said this week that OpenAI was “supposed to benefit humanity,” and now “it’s simply benefiting one of the richest companies in the world.”
2020-09-25 14:00:00+00:00 Read the full story…
Weighted Interest Score: 2.4521, Raw Interest Score: 1.2154,
Positive Sentiment: 0.5105, Negative Sentiment 0.0972
CloudQuant Thoughts : One minute we are lauding Microsoft for their AI Photo regeneration and Open Data Campaign, the next we have this…
CloudQuant Researchers prove Alpha in Tesseract Machine Learning Dataset
CloudQuant’s research team have published a white paper on the Tesseract data set which was constructed with a combination of cutting edge statistics and sophisticated machine learning methods. CloudQuant’s researchers concluded that the Tesseract Signal identifies both long and short investment signals that produce statistically significant investment return (alpha) at a greater than 99.9% (p-value < 0.001) level of confidence from 2018-2020. For information on this white paper and/or access to the data and back-test code used either… Email Sales@CloudQuant.com,
Make an appointment to speak to a CloudQuant Representative,
Or fill in the form on the right to be contacted back by a CloudQuant Representative.
Apple’s AI plan: a thousand small conveniences because AI is too dumb to do anything else
AI has become an integral part of every tech company’s pitch to consumers. Fail to hype up machine learning or neural networks when unveiling a new product, and you might as well be hawking hand-cranked calculators. This can lead to overpromising. But judging by its recent WWDC performance, Apple has adopted a smarter and quieter approach.
Why blind them with science when you can charm them with convenience? Sprinkled throughout Apple’s announcements about iOS, iPadOS, and macOS were a number of features and updates that have machine learning at their heart. Some weren’t announced onstage, and some features that almost certainly use AI weren’t identified as such, but here’s a quick recap of the more prominent mentions that we spotted:
- Facial recognition for HomeKit. HomeKit-enabled smart cameras will use photos you’ve tagged on your phone to identify who’s at your door and even announce them by name.
- Native sleep tracking for the Apple Watch. This uses machine learning to classify your movements and detect when you’re sleeping. The same mechanism also allows the Apple Watch to track new activities like dancing and…
- Handwashing. The Apple Watch not only detects the motion but also the sound of handwashing, starting a countdown timer to make sure you’re washing for as long as needed.
- App Library suggestions. A folder in the new App Library layout will use “on-device intelligence” to show apps you’re “likely to need next.” It’s small but potentially useful.
- Translate app. This works completely offline, thanks to on-device machine learning. It detects the languages being spoken and can even do live translations of conversations.
- Sound alerts in iOS 14. This accessibility feature wasn’t mentioned onstage, but it will let your iPhone listen for things like doorbells, sirens, dogs barking, or babies crying.
- Handwriting recognition for iPad. This wasn’t specifically identified as an AI-powered feature, but we’d bet dollars to donuts it is. AI is fantastic at image recognition tasks, and identifying both Chinese and English characters is a fitting challenge.
2020-06-25 00:00:00 Read the full story…
CloudQuant Thoughts : Let’s face it, Siri is the closest most people come to knowingly interacting with AI. Apple focusing on showing you how their AI is making your life simpler and more convenient definitely improves the public’s perception of AI.
British Startup Develops The AI-Accelerated Computational Alice Camera
The British startup, Photogram AI has recently announced a new AI-powered camera called — Alice Camera. According to the company website, the camera — Alice is an “AI-accelerated computational camera” that has been designed to deliver better connectivity than a regular or advanced DSLR.
With global pandemic, and subsequent economic lockdown in hand, the company believes that video streaming has gone utterly mainstream. Along with that, smartphones have also been making considerable advancements in the field of computational photography. And that’s why Alice is trying to bridge that gap by bringing computational photography into DLSRs too for professional content creators.
According to the company’s website, Alice Camera is an interchangeable lens camera that is equipped with a dedicated AI chip which will “elevate machine learning and pushes the boundaries of what a camera can do.”
2020-09-23 08:23:11+00:00 Read the full story…
Weighted Interest Score: 2.9257, Raw Interest Score: 1.1325,
Positive Sentiment: 0.4153, Negative Sentiment 0.1888
CloudQuant Thoughts : Google have demostrated what dramatic improvements AI and ML and bring to cell phone photos. Nice to see it applied to the much maligned stand alone Camera. We need some interesting innovative tech!
Jim Cramer on Palantir going public at a $22 billion valuation (4 min Video)
After 17 years on the private market, data analytics company Palantir is making its public market debut. CNBC’s Jim Cramer and David Faber discuss how investors might react.
2020-09-25 00:00:00 Read the full story…
Weighted Interest Score: 3.3755, Raw Interest Score: 1.6878,
Positive Sentiment: 0.0000, Negative Sentiment 0.0000
CloudQuant Thoughts : I think we all have our own opinions of Palantir, even their IPO is ‘unusual’.
5? is the magic number
5 Reasons Python is Still the King of Programming Languages
Just about every programming language has an ardent fanbase, and Python is no different. Long an extremely popular “generalist” language, Python has been establishing new fans in ultra-specialist segments such as data science and machine learning. No wonder it regularly ranks so highly on various “most popular language” lists, including the TIOBE Index, RedMonk, and Stack Overflow’s annual Developer Survey.
If you’re new to programming and wondering whether to prioritize the time to learn Python, here’s a brief run-through of what developers and other technologists love about the language, along with some advice about adopting it.
- It’s Easy to Learn
- Less Coding
- There’s a Massive Community and Tons of Add-ons
- Python is Growing
- Python Gets You Hired
2020-09-24 00:00:00 Read the full story…
Weighted Interest Score: 2.7787, Raw Interest Score: 1.8209,
Positive Sentiment: 0.3807, Negative Sentiment 0.0993
Five Key Features for a Machine Learning Platform
Machine learning platform designers need to meet current challenges and plan for future workloads.
As machine learning gains a foothold in more and more companies, teams are struggling with the intricacies of managing the machine learning lifecycle.
The typical starting point is to give each data scientist a Jupyter notebook backed by a GPU instance in the cloud and to have a separate team manage deployment and serving, but this approach breaks down as the complexity of the applications and the number of deployments grow.
As a result, more teams are looking for machine learning platforms. Several startups and cloud providers are beginning to offer end-to-end machine learning platforms, including AWS (SageMaker), Azure (Machine Learning Studio), Databricks (MLflow), Google (Cloud AI Platform), and others. Many other companies choose to build their own, including Uber (Michelangelo), Airbnb (BigHead), Facebook (FBLearner), Netflix, and Apple (Overton).
Weighted Interest Score: 2.4708, Raw Interest Score: 1.6981,
Positive Sentiment: 0.1992, Negative Sentiment 0.0734
Five Ways to Drive ROI with AI
AI is often touted as the way of the future for enterprises in all industries – but ensuring that the return on investment (ROI) from an AI implementation actually comes to fruition can often be a trickier thing. A group of AI-oriented companies – Appen, Cognizant, Cortex, Dataiku, DataRobot (which recently commissioned its own ROI study), and Deloitte – partnered to commission a study from ESI ThoughtLab that benchmarked 1,200 organizations to identify the factors that drive ROI from AI. The result: a roadmap for success in enterprise AI.
In addition to data on AI investments and returns, the cross-industry survey collected detailed data on how and why the 1,200 organizations had implemented AI. Using that data, combined with an AI maturity framework, input from an advisory board of AI experts, and in-depth interviews with AI leaders, ESI ThoughtLab arrived at a series of conclusions about the current state of AI in business.
- Begin with pilots, but then scale AI across the enterprise.
- Lay a firm foundation.
- Get your data right.
- Solve the human side of the equation.
- Adopt a culture of collaboration and learning.
2020-09-24 00:00:00 Read the full story…
Weighted Interest Score: 6.5024, Raw Interest Score: 2.0935,
Positive Sentiment: 0.3441, Negative Sentiment 0.2007
ICIJ Turns to Big Data Tech to Unravel FinCEN Files
Unraveling financial crimes like money laundering is a notoriously difficult task, especially when criminals purposely cover their tracks. It gets a little easier when you have advanced tools, such as text analytics, machine learning, and a graph database, which is what the International Consortium of Investigative Journalists (ICIJ) used with its latest investigation, dubbed the FinCEN Files.
The FinCEN Files is based on the leak of about 2,100 suspicious activity reports, known as SARS, that were sent to the U.S. Treasury’s Financial Crimes Enforcement Network, or FinCEN, between 2011 and 2017. SARS are written by compliance officials at banks in the US whenever fraud is suspected in a transaction. Anytime a transaction involves US dollars, it must go through a US bank, which means they pile up at the US Treasury Department.
2020-09-25 00:00:00 Read the full story…
Weighted Interest Score: 1.8951, Raw Interest Score: 1.0144,
Positive Sentiment: 0.1252, Negative Sentiment 0.3507
How Money Laundering Concerns Require New AI Monitoring Solutions
As money laundering capabilities evolve and become more complex, the financial sector is meeting those challenges with new AI monitoring solutions.
Artificial intelligence has created a number of amazing opportunities for the financial sector. The benefits of AI are endless. Financial institutions are using AI to enhance decision-making, improve customer service, project customer needs and much more. We have talked about the benefits of using big data and AI to improve cybersecurity. But there are other processes that could be equally important for financial institutions. AI can solve some pressing challenges that financial institutions can’t afford to overlook. This includes the growing threat of money laundering.
2020-09-27 20:43:23+00:00 Read the full story…
Weighted Interest Score: 3.1689, Raw Interest Score: 1.6142,
Positive Sentiment: 0.1932, Negative Sentiment 0.8002
Google Announces General Availability Of AI Platform Prediction
Gecently, the developers at Google Cloud announced the general availability of the AI Platform Prediction. The platform is based on a Google Kubernetes Engine (GKE) backend and is said to provide an enterprise-ready platform for hosting all the transformative ML models.
Emerging technologies like machine learning and AI have transformed the way most processes and industries work around us. Machine learning has brought various significant features that require predictions, such as identifying objects in images, recommending products, optimising market campaigns and more.
However, building a robust and enterprise-ready machine learning environment can include various issues like it being time-consuming, costly as well as complex. Google’s AI Platform Prediction takes into account all these issues to provide a robust environment for ML-based tasks.
2020-09-28 12:30:39+00:00 Read the full story…
Weighted Interest Score: 5.5645, Raw Interest Score: 2.6908,
Positive Sentiment: 0.1004, Negative Sentiment 0.0201
Google launches AI Platform Prediction in general availability
Google today launched AI Platform Prediction in general availability, a service that lets developers prep, build, run, and share machine learning models in the cloud. It’s based on a Google Kubernetes Engine backend and features an architecture designed for high reliability, flexibility, and low overhead latency.
IDC predicts that worldwide spending on cognitive and AI systems will reach $77.6 billion in 2022, up from $24 billion in revenue last year. Gartner agrees: In a recent survey of executives from thousands of businesses worldwide, it found that AI implementation grew a whopping 270% in the past four years and 37% in the past year alone. With AI Platform Prediction, Google adds yet another managed AI service to its portfolio, beating back competitors like Amazon, Microsoft, and IBM.
2020-09-25 00:00:00 Read the full story…
Weighted Interest Score: 5.5636, Raw Interest Score: 2.5307,
Positive Sentiment: 0.2169, Negative Sentiment 0.0362
How to make your trading firm data driven – Cuemacro
The question I get asked most, usually revolves around burgers, a variation of where’s my favourite burger place, or what cheese works best on a burger etc. I am always happy to answer these crucial questions!
Another question which I get asked often, is how do I begin to use data in my trading? The question is kind of easy, right? However, the answer is pretty involved, and it really depends on your firm. What might be appropriate for a quant firm, isn’t necessarily right for a discretionary firm, and you must tread carefully. Furthermore, everyone’s investment mandate is slightly different. In The Book of Alternative Data, Alexander Denev and I discuss how you can create teams to use alternative data. Below, I talk about some of these points from the book and also expand upon them to answer the more general question of how you can transform your trading firm to use data better.
2020-09-26 00:00:00 Read the full story…
Weighted Interest Score: 5.0899, Raw Interest Score: 1.7044,
Positive Sentiment: 0.2206, Negative Sentiment 0.3409
Education Provider BigBrainBank Launches AI Trading Signals
Asian fintech-edutech provider BigBrainBank.org has announced the launch of a new app called “TheBrain” to provide customers with trading signals using AI-powered analytics and algorithmic trend monitoring. According to the company’s CEO, Brendon Yong, the app TheBrain – AI Trade Strategies is available on both Android and iOS app stores as well as desktop version. The app significantly helps both corporate and retail investors to improve their trading result through trade ideas, risk on risk off, backtester and a host of advanced features.
The company indicated that “the algorithmic system in BigBrainBank.org also offers users extraordinary user interface and seamless navigation to extract important metrics in trend analysis and identify trade ideas with higher success rates and risk management measures in place.”
2020-09-22 13:53:11+00:00 Read the full story…
Weighted Interest Score: 4.5989, Raw Interest Score: 2.1494,
Positive Sentiment: 0.1535, Negative Sentiment 0.1024
First Industrywide Graph DB Conference Set for Sept. 28-30
Graph algorithms are the driving force behind the next generation of AI and machine learning that will power more and more industries and use cases as time goes on. Here is an opportunity to learn about how this all works. Event host TigerGraph, which makes a graph database that it claims is the only scalable one available for enterprises, has announced the final agenda and speaker lineup for “Graph + AI World 2020,” the first industry conference devoted to democratizing and accelerating artificial intelligence and machine learning through graph algorithms and graph analytics. The online event will be held Sept. 28-30.
More than 3,000 registrants are expected to attend the free virtual event, including data scientists, data engineers, architects, and business and IT executives from more than 100 companies from the Fortune 500. The final roster includes speakers from UnitedHealth Group, Intel, JPMorgan Chase, Jaguar Land Rover, Intuit, AT&T, Xandr (part of AT&T), Scotiabank, Accenture, KPMG, Publicis Sapient, Xilinx, eWEEK, and innovative startups that include Near, Ippen Digital, OpenCorporates, Expero, Abhay Solutions, SaH Analytics International, CAS, FinTell and Landing.AI.
2020-09-25 00:00:00 Read the full story…
Weighted Interest Score: 4.4212, Raw Interest Score: 1.8942,
Positive Sentiment: 0.1973, Negative Sentiment 0.1973
Webinar: Lending digitalisation & AI: Beyond the hype – what’s concrete?
The coronavirus crisis has escalated the need for financial institutions to digitise their processes.
One specific area that’s come under the spotlight is lending. The digitalisation of a financial institution’s lending process is no longer an option, but a requirement in this current economic climate. Firms must consider the compelling benefits of artificial intelligence (AI) when digitalising their credit process in order to overcome the COVID-19 economic crisis and stay ahead of competition.
In this free webinar, axefinance experts will showcase real customers’ use cases and will demonstrate the benefits of AI throughout the entire credit lifecycle, from onboarding to decision making.
2020-09-25 16:00:09+00:00 Read the full story…
Weighted Interest Score: 4.2315, Raw Interest Score: 1.8680,
Positive Sentiment: 0.2491, Negative Sentiment 0.2491
3 Ways to Build Neural Networks in TensorFlow with the Keras API
Building Deep Learning models with Keras in TensorFlow 2.x is possible with the Sequential API, the Functional API, and Model Subclassing
If you are going around, checking out different tutorials, doing Google searches, spending a lot of time on Stack Overflow about TensorFlow, you might have…
2020-09-28 07:41:48.292000+00:00 Read the full story…
Weighted Interest Score: 4.1192, Raw Interest Score: 2.1000,
Positive Sentiment: 0.2179, Negative Sentiment 0.0892
Hands-On Guide To Using AutoNLP For Automating Sentiment Analysis
Automated Machine learning or autoML is used for automating the complete process of machine learning for real-world problems to make the process easier and more efficient. Over the years researchers have developed ways of automating processes by developing tools like AutoKeras, AutoSklearn and even no-coding platforms like WEKA and H2o.
One such area of automation is in the field of natural language processing. With the d…
2020-09-28 07:30:03+00:00 Read the full story…
Weighted Interest Score: 3.7275, Raw Interest Score: 1.5595,
Positive Sentiment: 0.2350, Negative Sentiment 0.0427
Build The Next Best Code Curator With MachineHack’s New Hackathon
“Can you come up with an algorithm that can predict the bugs, features, and questions based on GitHub titles?”
An average smartphone OS contains more than 10 million lines of code. A million lines of code take 18000 pages to print which is equal to Tolstoy’s War and Peace put together 14 times! There is always a simpler, shorter version of the code along with a longer more exhaustive version.
The number of tools, languages, techniques, and appl…
2020-09-28 04:30:36+00:00 Read the full story…
Weighted Interest Score: 3.7050, Raw Interest Score: 1.8202,
Positive Sentiment: 0.2184, Negative Sentiment 0.3640
Modernizing Data Architectures for a Digital Age Using Data Virtualization
There’s a wide range of reasons why many organizations are deciding to modernize their data architectures. But they all agree on one thing: by using data more effectively, more widely, and more deeply, they can improve and optimize business and decision-making processes that will help them stay competitive in the emerging digital economy.
To prepare data architectures for the next evolution of analytics, the current systems that rely on physical…
2020-09-21 00:00:00 Read the full story…
Weighted Interest Score: 3.6055, Raw Interest Score: 1.9088,
Positive Sentiment: 0.2121, Negative Sentiment 0.2121
dotData Enterprise is Now Available on Microsoft Azure
dotData, focused on delivering full-cycle data science automation and operationalization for the enterprise, is providing dotData Enterprise on Microsoft Azure, offering increased speed and efficiency of data science and machine learning processes coupled with Azure’s strong managed IaaS/PaaS capabilities.
In addition, dotData has added Microsoft SQL Server and Azure SQL Database connectors which allows users to quickly and easily develop AI/ML models based on data stored in their corporate databases.
2020-09-25 00:00:00 Read the full story…
Weighted Interest Score: 3.5919, Raw Interest Score: 1.7604,
Positive Sentiment: 0.3444, Negative Sentiment 0.0000
Microsoft teams up with OpenAI to exclusively license GPT-3 language model
One of the most gratifying parts of my job at Microsoft is being able to witness and influence the intersection of technological progress and impact: harnessing the big trends in computing that have the opportunity to benefit everybody on the planet. Frank’s post this morning from Ignite shows just how much progress is happening in many of these areas.
Today, the foremost computing trend is undoubtedly artificial intelligence (AI). As we increasingly develop the ability to deploy huge AI models at scale in a way that can be leveraged by all developers and businesses, AI is becoming a platform – an environment upon which folks can build amazing new experiences, just like we’ve seen happen before with personal computers, mobile devices or the internet.
Getting this AI platform off the ground requires unprecedented computing horsepower. So, this May, we expanded upon our ongoing partnership with the world-leading AI research organization OpenAI to announce one of the world’s most powerful supercomputers – a custom-designed, Azure-hosted home for training OpenAI’s equally massive AI models.
2020-09-22 00:00:00 Read the full story…
Weighted Interest Score: 3.5625, Raw Interest Score: 1.6147,
Positive Sentiment: 0.5568, Negative Sentiment 0.0000
Amazon vets raise $4M from Madrona, Bezos Expeditions, others for AI2 spinout WhyLabs
Company leaders know they need to implement artificial intelligence and machine learning technologies within their businesses to stay ahead of the competition. But studies show that most organizations aren’t yet seeing an impact from AI investments and are increasingly wary of potential risks related to the burgeoning tech.
WhyLabs wants to help. The Seattle startup came out of stealth mode this week, unveiling its AI data monitoring platform that has attracted interest from top investment firms. Madrona Venture Group, Defy Partners, Bezos Expeditions — the VC arm of Amazon CEO Jeff Bezos — and Ascend VC participated in a $4 million round for the company, which is the latest to spin out of Seattle’s Allen Institute for Artificial Intelligence (AI2).
2020-09-23 14:34:00+00:00 Read the full story…
Weighted Interest Score: 3.4933, Raw Interest Score: 1.6334,
Positive Sentiment: 0.1199, Negative Sentiment 0.2398
Data Lake Vs. Data Warehouse: What Is The Difference?
When comparing data lake vs. data warehouse, it’s important to know that these two things actually serve quite different roles. They manage data differently and serve their own types of functions.
The market for data warehouses is booming. One study forecasts that the market will be worth $23.8 billion by 2030. Demand is growing at an annual pace of 29%.
While there is a lot of discussion about the merits of data warehouses, not enough discussion centers around data lakes. We talked about enterprise data warehouses in the past, so let’s contrast them with data lakes.
Both data warehouses and data lakes are used when storing big data. On the other hand, they are not the same. A data warehouse is a storage area for filtered, structured data that has been processed already for a particular use, while Data Lake is a massive pool of raw data and the aim is still unknown.
2020-09-23 20:41:11+00:00 Read the full story…
Weighted Interest Score: 3.4828, Raw Interest Score: 1.9173,
Positive Sentiment: 0.1407, Negative Sentiment 0.0704
Harnessing alternative data in the fight against fraud
The recent global crisis has set off a major fraud resurgence. With the world continuing its acceleration towards becoming digital-first, and with everything from work and transactions to entertainment and shopping happening online, potential attack vectors and opportunities are exponentially growing. The UK alone has seen a 66 percent rise in scams during the pandemic so far.
This is especially true for the financial services sector, as banks and financial organisations quickly shift their operations online during the pandemic in order to reach newly remote customers. Actions such as onboarding and sensitive transactions have been forced to take place purely remotely, while ID verification methods had to be adapted to cater to remote customers – during lockdown, the FCA even announced plans to accept selfies as part of a holistic identity verification process.
Fighting fraud with traditional techniques is no longer enough. As fraud becomes digital-first, so should anti-fraud techniques – businesses need to combine technology and data to create intelligent, real-time responses to problems, without a customer, or potential fraudster, ever even knowing. To do this, alternative data and machine learning are quickly becoming go-to solutions.
2020-09-28 00:00:00 Read the full story…
Weighted Interest Score: 3.4262, Raw Interest Score: 1.4047,
Positive Sentiment: 0.1155, Negative Sentiment 0.7697
TIBCO Introduces New Platforms to Better Manage Data
TIBCO Software Inc., a provider of enterprise data solutions, is focusing on disrupting the analytics space with a series of platform releases, including TIBCO Hyperconverged Analytics, providing immersive, smart, and real-time analytics to data-driven businesses. The company also unveiled TIBCO Spotfire 11 and TIBCO Cloud Data Streams, accelerating insights and actions for businesses.2020-09-24 00:00:00 Read the full story…
Weighted Interest Score: 3.1368, Raw Interest Score: 1.9657,
Positive Sentiment: 0.0836, Negative Sentiment 0.0418
Bidgely Secures US$8 Million Financing from CIBC Innovation Banking to Accelerate Growth
CIBC Innovation Banking is pleased to announce a US$8 million growth capital financing for cloud-based energy analytics software provider, Bidgely. Based in Mountain View, California, this funding will enable Bidgely to accelerate its growth plans across Asia, Europe and North America.
Bidgely uses artificial intelligence (AI) solutions to transform utility meter data into business intelligence for utilities and energy retailers who seek to better understand their customers. Retailers can leverage the augmented insights to personalize acquisition strategies and customer engagement, optimize retention, and to modernize grid operations. Bidgely works with over 25 utilities and energy retailers across the globe, and …
2020-09-24 00:00:00 Read the full story…
Weighted Interest Score: 2.9863, Raw Interest Score: 1.4263,
Positive Sentiment: 0.4037, Negative Sentiment 0.0000
Rethinking The Way We Benchmark Machine Learning Models
“Unless you have confidence in the ruler’s reliability, if you use a ruler to measure a table, you may also be using the table to measure the ruler.” Wittgenstein’s ruler
Do machine learning researchers solve something huge every time they hit the benchmark? If not, then why do we have these benchmarks? Benchmarks indeed guide researchers and their research objectives. But, if the benchmark is breached every couple of months then research objectives might become more about chasing benchmarks than solving bigger problems.
In order to address these challenges, researchers at Facebook AI have introduced Dynabench, a new platform for dynamic data collection and benchmarking. Dynabench can be used to collect human-in-the-loop data dynamically, against the current state-of-the-art, in a way that more accurately measures progress.
What’s Wrong With Current Benchmarks
Benchmarks are meant to challenge the ML community for longer durations. The rate at which AI expands can make existing benchmarks saturate quickly. With a new NLP model being released almost every two months, benchmarks fall back.
2020-09-28 09:30:00+00:00 Read the full story…
Weighted Interest Score: 2.7209, Raw Interest Score: 1.2631,
Positive Sentiment: 0.2831, Negative Sentiment 0.4138
Qeexo Adds Support for Arm’s Edge Processor
Qeexo, the “tinyML” specialist, said its AutoML platform now supports the smallest Cortex processors from Arm Ltd., making it the first vendor to automate machine learning on the Arm processor used for edge computing in sensors and microcontrollers.
The Carnegie Mellon University spinoff said Wednesday (Sept. 23) its AutoML platform that migrated to the cloud this past summer supports Arm’s Cortex-MO and -MO+ architectures aimed at edge computing applications. The “plus” version further reduces power consumption, a critical requirement for Internet of Things sensors and other unattended devices.
The Cortex-MO product line targets embedded applications and smart, connected devices used in industrial, automotive and other edge deployments.
2020-09-23 00:00:00 Read the full story…
Weighted Interest Score: 2.6676, Raw Interest Score: 1.9913,
Positive Sentiment: 0.1086, Negative Sentiment 0.0724
How Cloudera Enables Enterprises to Address Radical Change
Discussions of leading cloud computing often focus on the handful of U.S.-based companies–AWS, Microsoft Azure, IBM and Google–that lead the industry in terms of market share. That makes sense on one level but tends to obscure numerous other vendors, whose assistance is crucial to enterprises determined to capture the greatest value from their cloud computing and related investments.
One of the key players in this space is Cloudera. Founded in 2008, the company was an early mover in big data platforms and applications. However, Cloudera has also evolved steadily through organic development, acquisitions and strategic partnerships with key enterprise and cloud vendors to become a trusted partner for organizations of every kind.
Last week, the company announced new and upcoming data services based on the Cloudera Data Platform (CDP). Coming a year after the company purchased Arcadia Data, a provider of cloud-native, AI-driven business intelligence and analytics solutions, makes that acquisition seem particularly prescient. Let’s consider these new offerings and what they say about Cloudera’s position in the rapidly evolving and growing market for enterprise cloud.
2020-09-23 00:00:00 Read the full story…
Weighted Interest Score: 2.6258, Raw Interest Score: 1.4715,
Positive Sentiment: 0.3165, Negative Sentiment 0.1108
The Top Trends in Data Management for 2021 (Webinar)
From the rise of hybrid and multicloud architectures, to the impact of machine learning and automation, the business of data management is constantly evolving with new technologies, strategies, challenges and opportunities. The demand for fast, wide-range access to information is growing. At the same time, the need to effectively integrate, govern, protect and analyze data is also intensifying. All the while, data environments are increasing in size and complexity — traversing relat…
2020-12-10 00:00:00 Read the full story…
Weighted Interest Score: 2.5974, Raw Interest Score: 1.6729,
Positive Sentiment: 0.0929, Negative Sentiment 0.0929
Allen Institute researchers find pervasive toxicity in popular language models
Researchers at the Allen Institute for AI have created a data set — RealToxicityPrompts — that attempts to elicit racist, sexist, or otherwise toxic responses from AI language models, as a way of measuring the models’ preferences for these responses. In experiments, they claim to have found that no current machine learning technique sufficiently protects against toxic outputs, underlining the need for better training sets and model architectures.
It’s well-established that models amplify the biases in data on which they were trained. That’s problematic in the language domain, because a portion of the data is often sourced from communities with pervasive gender, race, and religious prejudices. AI research firm OpenAI notes that this can lead to placing words like “naughty” or “sucked” near female pronouns and “Islam” near words like “terrorism.” Other studies, like one published by Intel, MIT, and Canadian AI initiative CIFAR researchers in April, have found high levels of stereotypical bias from some of the most popular models, including Google’s BERT and XLNet, OpenAI’s GPT-2, and Facebook’s RoBERTa.
2020-09-25 00:00:00 Read the full story…
Weighted Interest Score: 2.5496, Raw Interest Score: 1.4179,
Positive Sentiment: 0.0773, Negative Sentiment 0.2578
Modern Data Warehousing: Enterprise Must-Haves (Webinar)
To fit into modern analytics ecosystems, legacy data warehouses must evolve – both architecturally and technologically – to deliver the agility, scalability and flexibility that business need to thrive in today’s data-driven economy. Alongside new architectural approaches, a variety of technologies have emerge…
2020-11-19 00:00:00 Read the full story…
Weighted Interest Score: 2.5448, Raw Interest Score: 1.6053,
Positive Sentiment: 0.0944, Negative Sentiment 0.0000
Evolutionary Decision Trees: When Machine Learning draws its Inspiration from Biology
The Decision Tree shows that for business travels the main factor of customer satisfaction is the online boarding: an easy and efficient online boarding increases the likelihood for customers to be satisfied. It also highlights the importance of the quality of inflight service wifi.
As our knowledge in Biology, or the Science of Life, increased tremendously over time, it has become a great source of inspiration for many engineers seeking to address challenging problems and develop creative innovations.
Take the example of the Japanese high-speed train, Shinkansen, one of the fastest trains of the world, with speeds in the excess of 300km/h. During its conception, engineers encountered serious difficulties because of the massive amount of noise created by the displacement of air ahead of the trains, which can even cause structural damage to several tunnels. To address this issue, they turned to an unlikely source, the Kingfisher! This bird has an elongated beak that enables him to dive into the water to hunt with a minimal splash. Thus, by redesigning the train in the image of the bird, engineers were able not only to solve the initial issue but also reduce the trains’ electricity consumption by 15%, and increase the speed by 10%.
Using knowledge in Biology as a source of inspiration is also possible in Machine Learning. In this article, I will focus on one example: Evolutionary Decision Trees.
2020-09-27 13:21:05.278000+00:00 Read the full story…
Weighted Interest Score: 2.4537, Raw Interest Score: 1.2627,
Positive Sentiment: 0.3013, Negative Sentiment 0.1291
AI and IoT Applied to Supply Chains Are Driving Digital Twins
The combination of IoT and machine learning growing at the same time is leading to a rise in the use of digital twins in the supply chain, as a digital replica that can be used for various purposes. The connection with the physical model and the corresponding virtual model is established by generating real time data using sensors.
The Digital Twin Consortium, launched in August as a program of the Object Management Group, is working on defining a taxonomy and standards and enabling technology including AI and simulation. Engineers are being attracted to the work. Founding members include Ansys, Dell, GE, Lendlease, Microsoft and Northrop Grumman.
“IoT and ML are the raw materials and the tools—the insight is in the repository where we model processes and create context. While this might be a database or a data lake, the most interesting example of this for me is the digital twin,” wrote Scott Lundstrom, an analyst focused on the intersection of AI, IoT and Supply Chains, on his blog, Supply Chain Futures.
2020-09-24 20:42:42+00:00 Read the full story…
Weighted Interest Score: 2.2408, Raw Interest Score: 1.4676,
Positive Sentiment: 0.1910, Negative Sentiment 0.1005
Databases 101: Introduction to Databases for Data Scientists
Data science is one of the fast-growing fields that I can’t see slowing down any time soon. Not with how our data dependence is overgrowing day by day. Data science is all about data, collecting it, cleaning it, analyzing it, visualizing it, and using it to make our life better. Handling large amounts of data can be a challenging task for data scientists. Most of the time, that data we need to process and analyze is much larger than the capacity of our devices (the size of the RAM). Storing the information on the hard-drive might cause our code to be much slower.
Not to mention that in order to make sense of the data, and to process it efficiently, we need to have this data ordered in some way. Here where databases come to play. A database is defined as a structured set of data held in a computer’s memory or on the cloud that is accessible in various ways. As a data scientist, you will need to design, create, and interact with databases on most of the projects you will work on. Sometimes you will need to create everything from scratch, while at other times, you will just need to know how to communicate with an already existing database.
When I first started my journey in data science, handling databases was one of the most challenging aspects to master. That’s why I decide to write a series of articles about everything databases. This article will be a brief introduction to databases. What is SQL? Why do we need databases? And the different types of databases.
2020-09-27 21:56:47.481000+00:00 Read the full story…
Weighted Interest Score: 2.0019, Raw Interest Score: 1.1477,
Positive Sentiment: 0.1868, Negative Sentiment 0.1601
Seattle startup Attunely raises $6M to help debt collection agencies collect payment
Attunely is raising more cash to support increasing demand for its machine learning platform used by debt collection agencies. The Seattle startup raised a $6 million Series A round from Framework Venture Partners, Anthos Capital, Vulcan Capital, and others.
Founded in 2018 and spun out of Seattle-based startup studio Pioneer Square Labs, Attunely crunches data related to debt records and consumer interaction history, in addition to other information such as macroeconomic trends, to produce a “score” for consumers who owe payment.
Attunely CEO Scott Ferris said the company saw a surge in activity earlier this year from creditors and recovery agencies that are “seeking technology solutions to optimize revenue recovery on call center resource constraints.”
2020-09-24 14:00:00+00:00 Read the full story…
Weighted Interest Score: 1.9737, Raw Interest Score: 1.4286,
Positive Sentiment: 0.1099, Negative Sentiment 0.0549
Microsoft Launches Spatial Analytics, Other AI Services at Ignite
Microsoft rolled out an array of new AI services during its Ignite conference today, including Spatial Analysis, a new offering that uses computer vision algorithms to detect and count the number of people in a room.
Spatial Analysis, which is part of the Microsoft Azure Cognitive Service offering, can combine images from multiple cameras to count the number of people in a room. It can also understand the distances between them (handy for social distancing in the COVID-19 era), and figure out how long they’re waiting in line or standing in front of displays.
The technology has already been rolled out at RXR, a real estate company based in New York City that has embedded spatial analysis in its RxWell app to ensure occupants’ safety and wellness.
“When it came to developing RxWell, there was simply no other company that had the capability and the infrastructure to meet our comprehensive data, analytics, and security needs than Microsoft,” RXR CEO Scott Rechler says in a press release.
2020-09-22 00:00:00 Read the full story…
Weighted Interest Score: 1.9084, Raw Interest Score: 1.1464,
Positive Sentiment: 0.1764, Negative Sentiment 0.1470
Researcher Interview: Ziv Epstein, Research Associate, MIT Media Lab
Zivvy Epstein is a PhD student in the Human Dynamics group of the MIT Media Lab. His work integrates aspects of design and computational social science to model and understand cooperative systems. He focuses on new challenges and opportunities that emerge from a digital society, particularly in the domains of artificial intelligence and social media. His research centers around creating new technologies and insights that make the internet a better place. In a new study, Who gets credit for AI-generated art?, published in iScience, Epstein, his advisor Prof. David Rand, and their coauthors focused on how credit and responsibility should be allocated when AI is used to generate art.
2020-09-24 21:11:05+00:00 Read the full story…
Weighted Interest Score: 1.8716, Raw Interest Score: 0.7623,
Positive Sentiment: 0.1607, Negative Sentiment 0.3578
Scaling to Great Heights at the Ray Summit
If you haven’t yet heard about Ray, the open source Python framework for building distributed applications, then next week’s Ray Summit will provide a compelling introduction to what might be one of the cornerstone technologies of the next decade.
Ray emerged several years ago from UC Berkeley’s RISELab with a goal of radically simplifying the process of developing distributed applications. The software was designed to support any application written in any language, but in practice it’s been used mostly with machine learning applications written in Python.
What it does sounds almost too good to be true. Instead of hiring a large team of engineers or Kubernetes experts to get an application running in a distributed manner on a large cluster, a single developer can enable their application to run in a parallel manner with the addition of a few lines of code and about 30 minutes of work.
2020-09-24 00:00:00 Read the full story…
Weighted Interest Score: 1.8593, Raw Interest Score: 1.0508,
Positive Sentiment: 0.1911, Negative Sentiment 0.0382
This news clip post is produced algorithmically based upon CloudQuant’s list of sites and focus items we find interesting. We used natural language processing (NLP) to determine an interest score, and to calculate the sentiment of the linked article using the Loughran and McDonald Sentiment Word Lists.
If you would like to add your blog or website to our search crawler, please email firstname.lastname@example.org. We welcome all contributors.
This news clip and any CloudQuant comment is for information and illustrative purposes only. It is not, and should not be regarded as investment advice or as a recommendation regarding a course of action. This information is provided with the understanding that CloudQuant is not acting in a fiduciary or advisory capacity under any contract with you, or any applicable law or regulation. You are responsible to make your own independent decision with respect to any course of action based on the content of this post.