AI & Machine Learning News. 12, October 2020

AI and Machine Learning Newsletter

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?


Top 10 Announcements From NVIDIA GTC 2020 Event

Someday, trillions of AI devices and machines will populate the Earth – in homes, office buildings, warehouses, stores, farms, factories, hospitals, airports.” – Jensen Huang, Founder, NVIDIA

Today, NVIDIA kicks off its flagship event, NVIDIA GPU Technology Conference (GTC) 2020 with a number of significant announcements and updates. This year the conference will be held virtually, and like every year, it started with a keynote speech of Jensen Huang, CEO and founder of NVIDIA from his kitchen.

The announcements include updates and introduction of the various platforms, partnerships, among others that involve data centres, edge AI, healthcare and collaboration tools. Huang outlined the vision for “Age of AI” in his GTC keynote speech.

“AI requires a whole reinvention of computing – full-stack rethinking – from chips and systems to algorithms, tools and the ecosystem,” Huang said, standing in front of the stove of his Silicon Valley home.

2020-10-05 Read the Full Story…

CloudQuant Thoughts : We are big fans of everything NVIDIA does, the new cards are a major step up in power. This keynote video demonstrates NVIDIA’s reach. Nice Stove! This video is Part 1 of 9 videos in the Keynote.

Inventing Virtual Meetings of Tomorrow with NVIDIA AI Research

New AI breakthroughs in NVIDIA Maxine, cloud-native video streaming AI SDK, slash bandwidth use while make it possible to re-animate faces, correct gaze and animate characters for immersive and engaging meetings.

2020-10-05 Read the Full Story…

CloudQuant Thoughts : THIS IS AMAZING!

New Report Highlights AI’s Role in Businesses’ Pandemic Resilience

During the pandemic, businesses have scrambled to avoid large layoffs and cutbacks by seizing any opportunity to shore up lost revenue from the world’s quieted economies. Now, analytics firm RELX has released its third annual Emerging Tech Executive Report, which applied three years of data to examine AI’s impact on businesses’ successes (or failures) during the pandemic, as well as general trends in enterprise use of AI technologies.

The report’s conclusions are drawn from AI-focused interviews with more than a thousand senior executives in the U.S. across eight industries: government, healthcare, legal services, insurance, science and medical fields, banking, and agriculture.

In general, the report finds strong growth in use of AI technologies: 81% of executives reported use of AI tech in their businesses, up 33% since 2018; 75% reported that their businesses offered AI training, up 29% since 2018; and 95% reported believing that U.S. companies should invest in the AI workforce through educational initiatives (up 3% since 2018).

2020-10-09 00:00:00 Read the full story…
Weighted Interest Score: 4.5026, Raw Interest Score: 1.6411,
Positive Sentiment: 0.1746, Negative Sentiment 0.4539

CloudQuant Thoughts : As per usual, those stats stand out. 81% of execs reported use of AI, 75% offer AI training, 95% believe companies should invest in AI educational initiatives for their workforce.

East and West: AI worries divided down regional lines

The perceived risk of artificial intelligence is linked to people’s location and profession, according to a new study, with those in the West typically much more concerned about the technology. In contrast, less than one in ten people in China say AI will be mostly harmful, and the majority believe it will be mostly helpful.

The University of Oxford’s Internet Institute analysed survey data from a risk poll of 154,195 participants living in 142 countries to assess their attitudes towards the development of AI over the coming decades for its report, Global Attitudes towards Artificial Intelligence (AI) & Automated Decision Making.

The findings vary significantly across regions. North Americans and Latin Americans are most skeptical about the benefits of AI, with more than 40 per cent believing AI will be harmful, whilst only 25 per cent of those living in South East Asia and just 11 per cent of those living in East Asia expressed similar concerns. One outlier on perceived harm is China, according to the study. Just nine per cent of respondents believe AI will be mostly harmful, with 59 per cent of respondents saying it will mostly be beneficial.

2020-10-11 23:39:26+00:00 Read the full story…
Weighted Interest Score: 3.8251, Raw Interest Score: 1.6419,
Positive Sentiment: 0.1173, Negative Sentiment 0.5082

CloudQuant Thoughts : This is no surprise from a society built on abandoning ones liberties to centralized control.

Gender Bias In the Driving Systems of AI Autonomous Cars

Here’s a topic that entails intense controversy, oftentimes sparking loud arguments and heated responses. Prepare yourself accordingly. Do you think that men are better drivers than women, or do you believe that women are better drivers than men?

Seems like most of us have an opinion on the matter, one way or another.

Stereotypically, men are often characterized as fierce drivers that have a take-no-prisoners attitude, while women supposedly are more forgiving and civil in their driving actions. Depending on how extreme you want to take these tropes, some would say that women shouldn’t be allowed on our roadways due to their timidity, while the same could be said that men should not be at the wheel due to their crazed pedal-to-the-metal predilection.

What do the stats say? According to the latest U.S. Department of Transportation data, based on their FARS or Fatality Analysis Reporting System, the number of males annually killed in car crashes is nearly twice that of the number of females killed in car crashes.

2020-10-08 22:21:29+00:00 Read the full story…
Weighted Interest Score: 2.8415, Raw Interest Score: 1.0498,
Positive Sentiment: 0.0542, Negative Sentiment 0.1676

CloudQuant Thoughts : Intriguing, the idea of pitting an AI trained entirely by monitoring female drivers against one trained by monitoring only male drivers!

CloudQuant Increases Liberator’s Speed & Reach

CloudQuant today rolled out a major update to their industry leading Liberator/Rosetta APIs.

This technical release provides Improved Performance, Error Feedback, and Column Level Filtering to the increasing number of CloudQuant clients using our external API, as well as providing a boost to CloudQuant’s research tools – CQ AI, CQ Mariner, and CQ Explorer.

Our API enables external users to seamlessly integrate Liberator’s Power and Speed into their own environments. CloudQuant’s Liberator and suite of Technological Products dramatically cut the time from data acquisition to profit!

2020-10-04 00:00:00 Read the full story…

CloudQuant Thoughts : Our Liberator API is at the center of our Data Fabric, it provides data to all of our clients and all of our products. As such we are constantly improving it. Liberator’s ability to deliver precise Alternative Data into any user’s platform in a simple and clear manner is a major driver of our goal of dramatically cutting the time it takes to go from Raw Data to Profit. Liberator has also qualified for a Benzinga Award Nomination!

How Geospatial Data Drives Insight for Bloomberg Users

Stockbrokers and other Wall Street professionals who use Bloomberg terminals are always on the lookout for an edge. Increasingly, that edge is coming in the form of geospatial data that describes the movement of people and goods – as well as natural events like hurricanes and viral pandemics — through space and time.

When US Air Force veteran Bobby Shackelton arrived at Bloomberg about five years ago, he found some pretty basic uses of geospatial data in the Bloomberg Terminal, that all-important source of news and data relied upon by thousands of individuals who make their living trading on information.

For example, commodity traders used geospatial data to understand the number of oil tankers under sail in the ocean at any point in time, which describes the short-term supply of that crucial commodity. But aside from a few niche cases like that, there really wasn’t much going on with geospatial data at Bloomberg.

So Shackelton, who helped set up radar installations in the Air Force before building geospatial data solutions at what would become S&P Global Market Intelligence, set out to build a full-stack mapping solution for the Bloomberg Terminal. That offering, called MAP GO, quietly debuted about two years ago.

2020-10-05 00:00:00 Read the full story…
Weighted Interest Score: 3.5040, Raw Interest Score: 1.6230,
Positive Sentiment: 0.1185, Negative Sentiment 0.0829

How to spot a data charlatan

When data are too scarce to split, only a data charlatan tries to follow inspiration with rigor, peddling hindsight by mathematically rediscovering phenomena that they already know to be in the data and calling their surprise statistically significant. This distinguishes them from the open-minded analyst who deals in inspiration and the meticulous statistician who offers proof of foresight.
When data are plentiful, get in the habit of data-splitting so you can have the best of both worlds without cheating! Be sure to do analytics and statistics separately on separate subsets of your original pile of data.

  • Analysts offer you open-minded inspiration.
  • Statisticians offer you rigorous testing.
  • Charlatans offer you twisted hindsight that pretends to be analytics plus statistics.

2020-10-10 16:29:07.966000+00:00 Read the full story…
Weighted Interest Score: 2.4656, Raw Interest Score: 1.0812,
Positive Sentiment: 0.3041, Negative Sentiment 0.2618

The Next Industry Shift for Machine Learning Performance Enhancement

The Next Industry Shift for Machine Learning Performance Enhancement. It’s simpler than you think.

“If you don’t like the road you’re walking, start paving another one.” – Dolly Parton

Thousands of companies around the world, from small startups to global corporations, find great value in improving the performance of their supervised or unsupervised ML models, whether it’s a sales or demand forecast, a market basket analysis recommender, a customer classifier, a sales optimizer, a chatbot, an algorithmic trading pipeline, a document labeler, an elections forecast, a spam filter, a medical diagnosis solution, a route optimizer, a face recognizer or a self-driving car. And I’m not even going to get started on IoT.

However, all of them seem to attempt to increase accuracy (reduce error) by focusing on mainly two things:

  • Feature engineering (getting the most out of your features by crunching your dataset to death)
  • Model/parameter optimization (choosing the best model and best parameters even if you have to come up with a hybrid of several algorithms and iterate to infinity)

Both of the above are very necessary indeed, but there is a third process that adds value in a complementary way, which has traditionally been wildly underused in most data science projects and is now starting to take off.

Adding external data. Over 90% of the world’s data has been created in the last two years alone, and volumes are expected to continue growing exponentially. Every 6 hours, one quintillion bytes of data are generated globally. You can’t come up with an intuitive reference for how much that is without recurring to stars or atoms and still, that figure will seem laughable in a couple of years.

2020-10-12 02:36:42.188000+00:00 Read the full story…
Weighted Interest Score: 3.9216, Raw Interest Score: 1.7344,
Positive Sentiment: 0.2843, Negative Sentiment 0.1990

10 Best Machine Learning Courses in 2020

If you are ready to take your career in machine learning to the next level, then these top 10 Machine Learning Courses covering both practical and theoretical work will help you excel.

Practical/Hands-on Courses with Less Theory

  1. Practical Deep Learning for Coders FAST.AI Price: Free
  2. Code-First Introduction to Natural Language Processing by Fast.ai Price: Free
  3. Python for Data Science and Machine Learning Bootcamp Price: $129 (on sale $10-$20)
  4. DeepLearning.AI TensorFlow Developer Professional Certificate Price: $49/month
  5. Datacamp Data Science Path Price: $25/month or $300/year

Theoretical Courses with Less Practical work

  1. Machine Learning by Stanford University Price: $80
  2. Deep Learning Specialization Price: $49/month
  3. CS231n by Andrej Karpathy Price: Free
  4. Stat 451: Introduction to Machine Learning Price: Free
  5. MIT Introduction to Deep Learning | 6.S191 Price: Free

2020-10-10 00:00:00 Read the full story…
Weighted Interest Score: 4.3029, Raw Interest Score: 2.4533,
Positive Sentiment: 0.0976, Negative Sentiment 0.0697

How To Use DeepCognition To Build Drag And Drop Deep Learning Models Without Coding?

Deep learning is an integral part of artificial intelligence and the contributions done in the field is immense. With increasing research and development in deep learning, there has been an increase in the use of no-code platforms for deep learning as well. There are a lot of platforms that support machine learning and the processes like data visualization, processing etc. But there are few platforms that focus only on deep learning and one such platform is DeepCognition.

In this article, we will learn a little bit about DeepCognition and build an algorithm using DeepCognition platform.

Who are DeepCognition.ai? DeepCognition was founded with an aim of democratization of artificial intelligence. They have created a platform that can be used to create and deploy deep learning models with just clicking of buttons and no code at all. The problem they are trying to solve is to overcome the shortage of expertise in AI that is creating barriers in organizations in the adoption of AI and make deep learning accessible to all.

2020-10-12 08:30:53+00:00 Read the full story…
Weighted Interest Score: 3.9018, Raw Interest Score: 1.7634,
Positive Sentiment: 0.1206, Negative Sentiment 0.0904

Fractal Hives Off Theremin.ai After Raising Funds From OLMO Capital

Theremin.ai Raises Funds From OLMO Capital

In a recent LinkedIn post, Co-founder, Group Chief Executive & Vice-Chairman, Fractal Analytics, Srikanth Velamakanni, stated his excitement of sharing the news — “I am excited to share with you that theremin.ai, Fractal’s AI-driven automated investing business has raised funds from OLMO capital.”

He further stated that “We set up theremin.ai to test whether our algorithms could find signals in a nearly perfect capital markets context and we are encouraged by the results.”

2020-10-05 Read the Full Story…

Data Architecture and Artificial Intelligence: How Do They Work Together?

Artificial intelligence (AI) is rapidly gaining ground as core business competency. Today’s machine learning (ML) or deep learning (DL) algorithms promise to revolutionize business models and processes, restructure workforces, and transform data infrastructures to enhance process efficiency and improve decision-making throughout the enterprise. Gone are the days of data silos and manual algorithms.

However, widespread belief by stating that AI’s growth was stunted in the past mainly due to the unavailability of large data sets. Big data changed all that – enabling businesses to take advantage of high-volume and high-velocity data to train AI algorithms for business-process improvements and enhanced decision making.

The Road to AI Leads through Information Architecture describes howhybrid Data Management, Data Governance, and business analytics can together transform enterprise-wide decision making. According to this author, these three core business practices can enable organizations of all sizes “to unleash the power of AI in the enterprise.”

2020-09-29 Read the Full Story…

Machine learning for anomaly detection: Elliptic Envelope

Welcome back to anomaly detection; this is 6th in a series of “bite-sized” data science focusing on outlier detection. Today I am writing about a machine learning algorithm called EllipticEnvelope , which is yet another tool in data scientists’ toolbox for fraud/anomaly/outlier detection.
In case you have missed my previous articles or you are interested in learning more about the topic, find the links here (Local Outlier Factor (LOF), Z-score, Boxplot, Statistical techniques, Time series anomaly detection, Elliptical Envelope)
So what is elliptic envelope and what’s the intuition behind the algorithm? If you have taken geometry classes you are probably familiar with ellipse — a geometric configuration that takes an oval shape on a two-dimensional plane.

2020-10-06 Read the Full Story…

How AI Is Revolutionizing Social Visibility

Artificial intelligence has the power to revolutionize the social visibility of brands, making way for a very inclusive approach towards online marketing. Today, the power of digital marketing and artificial intelligence go hand in hand.

Artifical intelligence (AI) in digital marketing is useful in gathering data from all aspects, analyzing it, simplifying it, and then getting an easy understanding of a consumer’s needs and preferences.

The predictive assessment of social platforms is expected to grow to more than $2.1 billion in value by the year 2023. Hence, the market of predictive data, which utilizes the deep learning methods of AI, is burgeoning.

AI technology helps marketers gain insights that are more accurate and deliver customized consumer experiences. Development with AI has helped companies deal with customer interactions and analytics. This is why AI technology has all the capabilities to improve digital marketing strategies.

Why Organizations Value AI Technology to Improve Social Visibility:

2020-10-07 Read the Full Story…

Artificial Intelligence meets market volatility: Swiss tech firm opens hedge fund

The crisis is a good mean for revealing the relevancy of a successful investment strategy” says Vestun’s CEO as it opens its AI hedge fund to external retail investors

Vestun, a Swiss-based financial and technology company has now opened the launch of its hedge fund to new outside investors.

The firm which until now has been only managing its own capital announced that its investment vehicle will open to institutional investments including banks, multi-family offices and asset managers within certain jurisdiction.

The company flagship strategy trades liquid US equities systematically. The strategy is designed to autonomously adapts its portfolio and risk exposure dynamically to the prevailing market conditions. In contrast to traditional systematic strategies, Vestun’s approach does not rely on statistical rules and historical events to generate signals. Instead, the strategy aggregate domain specific intelligence with datasets that individually perform in their own economics while remaining uncorrelated against each other.

2020-10-07 Read the Full Story…

Top Twitter Accounts On AI One Must Follow

At the present scenario, Artificial Intelligence and machine learning have been portraying a critical role in the advancement of the tech sector. Social media platforms have been performing a significant role when it comes to keeping updated with the latest and trending information.

One such platform is Twitter. Twitter not only helps you keep track of the latest social and economic news, but also it allows you to both share and acquire knowledge about emerging technologies.

Below here, we jotted a list down the top ten AI accounts, based on alphabetical order, one must follow on Twitter.

2020-10-07 Read the Full Story…

India Is Working To Develop A Supercomputer To Facilitate AI Framework

India is working to develop a framework to help different walks of life in the longer-term in order to become an AI superpower. The Indian government is also working with the Centre for Development of Advanced Computing to develop a supercomputer to facilitate the AI framework.

AI has been widely believed to play a significant role in improving governance, along with some of the top use cases in the field of social welfare, policymaking, and healthcare. In fact, countries such as the US and China have already made giant strides in this direction. And thus India needed to make its way as well.

When asked, the CEO of National E-Governance Division, Abhishek Singh stated to the media that globally, we are recognised as a country which has a vast AI-skilled workforce, along with a good network of startup companies which are creating products. However, the only thing lacking is the compute capabilities, which is required. And that’s why the government is currently working on a framework and an ecosystem to facilitate that.

Singh further stated that computing facilities are being set up in India and will allow the AI, startups, tech entrepreneurs and researchers to leverage the infrastructure that has been built to run their algorithms and to create “world-class AI products.”

2020-10-05 Read the Full Story…

Sibos 2020: Victoria Harverson, global head of business development at SmartStream Air (Video)

FinTech Futures sat down with Victoria Harverson, global head of business development at SmartStream Air. Harverson was appointed into the role very recently, in order to run the business for SmartStream’s artificial intelligence (AI) solution for data processing.

Key highlights:

  • Developments that have taken place since SmarStream launched its first AI solution at last year’s Sibos.
  • What is on the agenda for the next 12 months.
  • Breaking down down how SmartStream Air works.
  • What it’s doing to transform traditional data verification and reconciliation processes.

2020-10-07 14:46:32+00:00 Read the full story…
Weighted Interest Score: 7.0922, Raw Interest Score: 3.2028,
Positive Sentiment: 0.0000, Negative Sentiment 0.0000

Sharp Venture Capitalists Make Remarkable Inroads With Alternative Data

Alternative data is not utilized nearly enough in data capital management. It can help you increase your returns in powerful ways.

The University of Hawaii reports that big data is shaking up the venture capital industry in unbelievable ways. Venture capitalists are finding new ways to leverage alternative data effectively for much higher yields.

Big data plays a role in shifting the risk-reward calculus in the favor of venture capitalists. Venture capital is a high risk, high reward game. To put it into perspective, 90% of new startups fail, which means that investors can lose a lot of money while hunting the potential “unicorns.” Historically, venture capital has been regarded more as an art form than a science.

Investors were known for following their intuitions, impressions, and carefully cultivated personal networks rather than relying on cold algorithms. This has changed in the era of big data, which is why investing apps that use data analytics have really taken off. Data capital management could be a huge thing in the future.

2020-10-08 08:34:08+00:00 Read the full story…
Weighted Interest Score: 5.9046, Raw Interest Score: 2.4377,
Positive Sentiment: 0.3750, Negative Sentiment 0.2625

Exabel makes CCO and data partnership hires

Exabel, an alternative data, AI and data science platform for active asset managers, has hired Jan Bratteberg as Chief Commercial Officer (CCO) and Nathaniel Cohn as VP of North America and Head of Data Partnerships.

The news follows the closure of a third funding round last month and the commercial launch of the firm earlier this year.

As CCO, Jan will build out further the commercial strategy of the company, as it seeks to disrupt the USD2 billion alternative data market by making the datasets available more usable by investors without the need for in-house quants or technology infrastructure. Jan joins from San Francisco based investment management firm Algert Global, where he has been a Partner since 2016. Prior to this, he was a Managing Director at BlackRock where he spent 17 years in a variety of investment roles.

2020-10-12 00:00:00 Read the full story…
2020-10-12 10:59:00 Read the full story…

Weighted Interest Score: 5.8135, Raw Interest Score: 2.0140,
Positive Sentiment: 0.1233, Negative Sentiment 0.2055

arXiv now allows researchers to submit code with their manuscripts

Papers with Code today announced that preprint paper archive arXiv will now allow researchers to submit code alongside research papers, giving computer scientists an easy way to analyze, scrutinize, or reproduce claims of state-of-the-art AI or novel advances in what’s possible.

An assessment of the AI industry released a week ago found that only 15% of papers submitted by researchers today publish their code.

Maintained by Cornell University, arXiv hosts manuscripts from fields like biology, mathematics, and physics, and it has become one of the most popular places online for artificial intelligence researchers to publicly share their work. Preprint repositories give researchers a way to share their work immediately, before undergoing what can be a long peer review process as practiced by reputable scholarly journals. Code shared on arXiv will be submitted through Papers with Code and can be found in a Code tab for each paper.
2020-10-08 00:00:00 Read the full story…
Weighted Interest Score: 4.9460, Raw Interest Score: 2.5997,
Positive Sentiment: 0.2426, Negative Sentiment 0.1733

6 Essential Skills To Be A Data Scientist

Leveraging the use of big data, as an insight-generating engine has driven the demand for data scientists at the enterprise-level, across all industry verticals. Whether it is to refine the process of product development, improve customer retention, or mine through the data to find new business opportunities — organizations are increasingly relying on the expertise of data scientists to sustain, grow, and outdo their competition.

Consequently, as the demand for data scientists increases, the discipline presents an enticing career path for students and existing professionals.

2020-10-12 00:10:34.359000+00:00 Read the full story…
Weighted Interest Score: 4.4840, Raw Interest Score: 2.4199,
Positive Sentiment: 0.3559, Negative Sentiment 0.1423

Practical Machine Learning Tutorial: Part.1 (Exploratory Data Analysis)

Although there are tons of great books and papers outside to practice machine learning, I always wanted to see something short, simple, and with a descriptive manuscript. I always wanted to see an example with an appropriate explanation of the procedure accompanied by detailed results interpretation. Model evaluation metrics should also need to be elaborated clearly.

In this work, I will try to include all important steps of ML modeling (even though some are not necessary for this dataset) to make a consistent and tangible example, especially for geoscientists. Eight important ML algorithms will be examined and results will be compared. I will try to have an argumentative model evaluation discussion. I will not go deep into the algorithm’s fundamentals.

To access the dataset and jupyter notebook find out my Git.
Note1: codes embedded in this manuscript are presented to understand the work procedure. If you want to exercise by yourself, I highly recommend using the jupyter notebook file.
Note2: shuffling data can cause differences between your runs and what appears here.

This tutorial has four parts:

  1. Exploratory Data Analysis,
  2. Build Model & Validate,
  3. Model Evaluation-1,
  4. Model Evaluation-2

2020-10-12 05:16:43.355000+00:00 Read the full story…
Weighted Interest Score: 4.1253, Raw Interest Score: 1.9341,
Positive Sentiment: 0.0784, Negative Sentiment 0.0784

Transfer Learning-Rock Paper Scissors Classifier

How to use transfer learning for classifying images.

Growing up building things using Lego has always been fun, so is building machine learning algorithm from scratch. Usually, machine learning algorithms are sufficient for various applications but when it comes to huge data size and classifying images we need more powerful machine learning algorithms hence deep learning comes into picture. Building an algorithm is always beneficial but time consuming so why not use existing algorithms and model for similar type of data. The process of using the stored knowledge which is gained while solving one problem and applying it to a different but similar problem is called Transfer Learning. Let’s get a better picture of how we can use some really powerful convolutional neural network on our own data set.
2020-10-11 19:13:15.264000+00:00 Read the full story…
Weighted Interest Score: 3.9540, Raw Interest Score: 1.8726,
Positive Sentiment: 0.1074, Negative Sentiment 0.0921

Develop and Deploy an Image Classifier App Using Fastai

Fastai is a popular open-source library used for learning and practicing machine learning and deep learning. Jeremy Howard and Rachel Thomas founded fast.ai with the objective of making deep learning more accessible. All the exhaustive resources such as courses, software, and research papers available in fast.ai are completely free.

In August 2020, fastai_v2 was released that promises to be much faster, and more flexible to implement deep learning frameworks. The 2020 fastai course combines the core concepts of both machine learning and deep learning. It also teaches the user about the important aspects of model production and deployment. In this article, I will discuss the techniques taught in the initial three lessons of the fast.ai beginner course, about building a quick and simple image classification model. Along with building the model, you will also learn how to easily develop a web application for the model and deploy it for production.

This article will follow the top-down approach that Jeremy follows for teaching in his courses. You will first learn about training an image classifier. Later, the details about the model used for classification will be explained. The prerequisite for understanding this article is knowledge of Python, as fastai is written in Python and built on PyTorch. It is recommended to run this code in Google Colab or Gradient, as GPU access is required. Also, fastai can be easily installed on these two platforms.

2020-10-08 09:48:48+00:00 Read the full story…
Weighted Interest Score: 3.4470, Raw Interest Score: 1.3642,
Positive Sentiment: 0.1441, Negative Sentiment 0.1249

Amplify Intelligence With AI And Analytics — Forrester’s Virtual Data & Insights Forum, October 13–15

They say data is the new oil. They say data is the new currency. They say data is the key competitive differentiator. All true. But reality is sobering: Only 7% of firms report advanced, insights-driven practices. Respondents to the same survey claimed that less than half (49%) of all business decisions in their enterprise are made based on quantitative information — a number that hasn’t moved much in the past three years (46% in 2018 and 48% in 2019). Lastly, anecdotal evidence shows that less than 20% of all raw business and operational data makes it into analytical databases and applications, and only 20% of enterprise knowledge workers who could be leveraging enterprise-grade analytical applications are doing so.

The reasons for the lack of more progress are plentiful and not new and are spread across the usual suspects: strategy, process, people, data, and technology. Our team — Business Insights — has plenty of research to help you move the needle and become more advanced in your insights-driven business capabilities. We plan to showcase much of that research during Forrester’s Data Strategy & Insights Forum on October 13–15. Specifically, the track that I have the privilege to lead — “Amplify Intelligence With AI and Analytics” — will showcase our latest research on how to scale AI and analytics across six sessions.

2020-10-05 15:15:40-04:00 Read the full story…
Weighted Interest Score: 3.4188, Raw Interest Score: 1.6606,
Positive Sentiment: 0.1215, Negative Sentiment 0.1620

NVIDIA Just Gave A PyTorch Based Conversational AI Model For Free

Last week, NVIDIA announced the NeMo model for the development of speech and language models and to create a conversational AI. NeMo is an open-source toolkit based on the PyTorch backend. The neural modules form the building blocks of these NeMo models. With NeMo, users can compose and train state-of-the-art neural network architectures.

How Can NeMo Help : NVIDIA NeMo allows to quickly build, train, and fine-tune conversational AI. It consists of NeMo core and NeMo collections. While NeMo core helps in getting the common look and feel for all models, NeMo collections act as groups of domain-specific modules and models.

There are main parts of NeMo: model, neural module, and neural type. The models contain all necessary information regarding training, fine-tuning, data augmentation, and infrastructure details.
2020-10-12 11:30:54+00:00 Read the full story…
Weighted Interest Score: 3.3703, Raw Interest Score: 1.7439,
Positive Sentiment: 0.0662, Negative Sentiment 0.0883

Register Now for Data Summit Connect Fall 2020

Registration is now open for Data Summit Connect Fall 2020, a series of data management and analytics webinars presented by DBTA and Big Data Quarterly, that will take place October 20-22.

In addition to the regular program, two pre-conference workshops, “Introduction To Knowledge Graphs” and “Getting Started With DataOps: Orchestrating The Three Pipelines,” will be offered on October 19.

Following on the success of Data Summit Connect in June, Data Summit Connect Fall 2020 will again provide practical advice, inspiring thought leadership, and in-depth training.

With travel plans still on hold and in-person meetings difficult, it is more important than ever to stay connected and in touch with peers and industry experts, and also to keep up-to-date with the latest technologies and industry trends.

2020-10-05 00:00:00 Read the full story…
Weighted Interest Score: 3.2810, Raw Interest Score: 1.7832,
Positive Sentiment: 0.2853, Negative Sentiment 0.2853

How Supercomputers Help To Create The Next Generation of Fully Integrated Data Centres

“Data centre is an asset that needs to be protected”- Michael Kagan, CTO of NVIDIA

On the first day of the NVIDIA GPU Technology Conference, Jensen Huang, founder of NVIDIA revealed the company’s three-year DPU roadmap that featured the new NVIDIA BlueField-2 family of DPUs and NVIDIA DOCA software development kit for building applications on DPU-accelerated data centre infrastructure services.

Michael Kagan, CTO of NVIDIA recently in a talk, explained the next generation of fully integrated data centres and how supercomputers and edge AI helps in augmenting such initiatives.

Kagan stated that the state-of-the-art technologies from both NVIDIA and Mellanox created a great opportunity to build a new class of computers, i.e. the fully-integrated cloud data centres that are designed to handle the workload of the 21st century.
2020-10-10 07:30:06+00:00 Read the full story…
Weighted Interest Score: 3.2201, Raw Interest Score: 1.7107,
Positive Sentiment: 0.1111, Negative Sentiment 0.2000

Raise 2020: Here Are The Top Quotes From The Event

For the last two decades, there have been two important developments. First, we have developed data on an exponential scale. Second, AI, Cloud Computing & Machine Learning has gained traction. Now it is our duty to use these developments which have matured in the last two decades for the good of our country.

The recently concluded RAISE 2020 – The Responsible AI for Social Empowerment virtual summit hosted a global meeting of best minds to exchange ideas around AI for social empowerment, inclusion and transformation in various industries. Inaugurated by Hon’ble Prime Minister Narendra Modi and attended by the likes of Mukesh Ambani, Ajay Sawhney, Ravi Shankar Prasad, Amitabh Kant, among others, the five-day event hosted many engaging sessions by these ingenious minds. We list some interesting statements made at RAISE 2020, in this article.

2020-10-12 10:30:25+00:00 Read the full story…
Weighted Interest Score: 3.2154, Raw Interest Score: 1.2945,
Positive Sentiment: 0.6472, Negative Sentiment 0.0000

Top 10 Ways AI Drives Price Optimization in Retail

There are several techniques in use in various stages if maturity in retail and e-commerce. Many different tools and techniques feed into AI powered price optimization for retailers. When used together these can drive very significant top line and bottom-line results for retailers, and allow them to be much more agile in their response to changes in market conditions like competition, costs, inventory levels, and more.

Here are the top 10 “need to know” concepts in the use of AI in price optimization for retailers.

  1. Segmentation of Customers and Products
  2. Regression or Stochastics Modeling
  3. Elasticity
  4. Dynamic Competitive Pricing
  5. Test and Learn
  6. Cannibalization
  7. Market Basket Optimization
  8. Promotion Optimization
  9. Recommendation Engines
  10. OCR, NLP

2020-10-06 00:00:00 Read the full story…
Weighted Interest Score: 3.1212, Raw Interest Score: 1.4741,
Positive Sentiment: 0.1504, Negative Sentiment 0.1354

Global Model Interpretability Techniques for Black Box Models

There is no mathematical equation for model interpretability. ‘Interpretability is the degree to which a human can consistently predict the model’s result’

An interpretable model that makes sense is far more trustworthy than an opaque one. There are two reasons for this. First, the business users do not make million-dollar decisions just because a computer said so. Second, the data scientists need interpretable models to ensure that no errors were made in data collection or modeling, which would otherwise cause the model to work well in evaluation, but fail miserably in production.

The importance of interpretability is subjective to the user of the model. The accuracy of a model may be more important than the interpretability of the model in cases where the model is used to power a solution. The data product is communicating with an entity or through an interface that eliminates the need for interpretability. However, when humans are the users of the model, interpretability takes a front seat.

2020-10-12 07:06:40+00:00 Read the full story…
Weighted Interest Score: 3.1067, Raw Interest Score: 1.7852,
Positive Sentiment: 0.0837, Negative Sentiment 0.2929

Finding new ways to operate & transform with machine learning (Video)

Mark Smith, Worldwide Head of Business and Market Development for Payments, Amazon Web Services gives his View From Sibos on the power of AI and machine learning. We learn about how compute power has become more accessible, giving companies the ability to harness the cloud and use AI&ML tools to tackle new issues, the impact COVID-19 has had on your customer’s journey to implementing machine learning workloads, and about the challenges financial institutions are still grappling with around machine learning.

2020-10-06 09:00:00 Read the full story…
Weighted Interest Score: 3.0476, Raw Interest Score: 1.9268,
Positive Sentiment: 0.0000, Negative Sentiment 0.1927

New Hackathon For Data Scientists – GitHub Bugs Prediction Challenge

MachineHack, in association with Embold, has recently launched a brand new hackathon — GitHub Bugs Prediction Challenge — where participants need to predict bugs on the GitHub titles and text body. The registration is now open and the hackathon closes on 18th of October 2020.

Embold.io is a software quality platform that enables leveraging quality code within a short duration. It combines machine learning, rigorous statistical algorithms, and powerful programming techniques to develop cutting edge products for the industry.

In this hackathon, data scientists need to come up with an algorithm that can predict the bugs, features, and questions based on GitHub text data. With this hackathon, participants will undergo an interesting learning curve where they will be able to write some quality code to win the prizes, as the evaluation involves getting a code quality score using the Embold Code Analysis platform. Further, Embold is also providing a quick tour of how to use its code analysis platform for free.

2020-10-08 09:07:04+00:00 Read the full story…
Weighted Interest Score: 3.0155, Raw Interest Score: 1.3957,
Positive Sentiment: 0.2641, Negative Sentiment 0.2263

Highlights from NVIDIA’s Landmark GPU Technology Conference

Artificial intelligence and chipmaker NVIDIA held its fall GPU Technology Conference (GTC) this week in an all-digital format. Like the live version, NVIDIA uses the event as a launch mechanism to articulate its vision and launch new products. The October 2020 GTC event was packed with announcements, partnerships, educational presentations and use cases.

The digital event let NVIDIA do things it could not do before. It’s the first GTC that ran across the world’s time zones, with sessions in Chinese, Korean, Japanese and Hebrew, all in local times. There were 1000 sessions this year, 400 more than last year.

There was a ton of content, but there were a few themes and newsworthy items that I thought stood out and they are below.

2020-10-08 00:00:00 Read the full story…
Weighted Interest Score: 2.9987, Raw Interest Score: 1.3368,
Positive Sentiment: 0.2119, Negative Sentiment 0.1141

Quick Guide to Evaluation Metrics for Supervised and Unsupervised Machine Learning

Machine learning is about building a predictive model using historical data to make predictions on new data where you do not have the answer to a particular question. Since, the output is probabilistic, evaluating your predictions becomes a crucial step. There are a lot of ways by which you can judge how well your machine learning model performs and mostly all of them focus on minimizing the error between the actual and predicted entity because you would want your predictions to be more and more accurate.

Supervised learning algorithms, where you have information about the labels like in classification, regression problems, and unsupervised learning algorithms, where you don’t have the label information such as clustering, have different evaluation metrics according to their outputs. In this post, you will explore some of the most popular evaluation metrics for classification, regression, and clustering problems. More specifically, you’ll :

  • learn all the terms related to the confusion matrix and metrics drawn from it
  • learn evaluation metrics like RMSE, MAE, R-Squared, etc. for regression problems
  • learn metrics like Silhouette coefficient, Dunn’s index for clustering problems

All the evaluation metrics described in this tutorial have an implementation available as libraries, packages on different platforms like Python, R, Spark, etc., however, this tutorial is only concerned with the meaning of these metrics which you should be aware of before using them. You can use this guide as a quick reference in case you need to quickly revise the important metrics in machine learning.

2020-10-12 08:45:08+00:00 Read the full story…
Weighted Interest Score: 2.9863, Raw Interest Score: 1.3575,
Positive Sentiment: 0.1211, Negative Sentiment 0.5608

AWS Cuts Prices for SageMaker GPU Instances

Amazon Web Services is cutting prices on its SageMaker managed service for machine learning and deep learning as it attracts more financial services, healthcare and retail customers building and training ML models in production.

The cloud giant (NASDAQ: AMZN) said Wednesday (Oct. 7) it is reducing prices for GPU instances running SageMaker by as much as 18 percent. Reminiscent of earlier price cuts as AWS battled Microsoft Azure and Google (NASDAQ: GOOGL) for public cloud dominance, the reductions for SageMaker reflect the growing number of enterprise options for building, training and deploying machine and deep learning as production workloads.

Released in late 2017, SageMaker was among the first model trainers out of the gate. Since then AWS has expanded the ecosystem to include tools for building and managing training data sets along with an integrated development environment dubbed SageMaker Studio. The IDE allows developers to collect and store code, notebooks, data sets, settings and project folders in a single place.

2020-10-07 00:00:00 Read the full story…
Weighted Interest Score: 2.7798, Raw Interest Score: 1.9331,
Positive Sentiment: 0.0000, Negative Sentiment 0.1487

Advanced Micro Devices Inc close to concluding takeover talks with Xilinx Inc

Advanced Micro Devices Inc (NASDAQ:AMD) is in talks about a potential US$30bn offer for semiconductor devices specialist Xilinx Inc (NASDAQ:XLNX), according to reports.

The move, first reported by the Wall Street Journal, comes amid fierce rivalry with chipmakers Intel Corporation (NASDAQ:INTC) and NVIDIA Corporation (NASDAQ:NVDA).

Xilinx, which has collaborated with AMD in the past, specialises in making field-programmable gate arrays (FPGA), in-demand semiconductor devices, and cutting-edge adaptive compute acceleration platform (ACAP) products, which both are used in data centers, wireless communications, AI and machine learning, electric cars, aerospace and defense.

The talks were reported to have resumed after a recent hiatus, but the WSJ said a decision could come as early as next week.

2020-10-09 00:00:00 Read the full story…
Weighted Interest Score: 2.6930, Raw Interest Score: 1.4388,
Positive Sentiment: 0.0899, Negative Sentiment 0.0899

Citi Develops ESG Platform To Transform Research

Citi is adding artificial intelligence-driven environmental, social, and governance data from Truvalue Labs to a proprietary platform that will allow the bank’s analysts to include the financial materiality of key ESG issues in research reports from the fourth quarter of this year.

Val Smith, Citi’s chief sustainability officer, said in a statement: “The Truvalue Labs collaboration with our Research & Global Insights team is an exciting development, as it will enable us to combine internal analysis, ESG data and AI to help us gain a deeper understanding of the opportunities and risk landscapes for our clients.”

Rich Webley, head of global data insights at Citi, told Markets Media that the bank selected TruValue after carrying out a detailed study over 12 months on how to use AI to incorporate ESG data.

2020-10-07 12:29:20+00:00 Read the full story…
Weighted Interest Score: 2.6467, Raw Interest Score: 1.6469,
Positive Sentiment: 0.1976, Negative Sentiment 0.0439

The secrets of small data: How machine learning finally reached the enterprise

Over the past decade, “big data” has become Silicon Valley’s biggest buzzword. When they’re trained on mind-numbingly large data sets, machine learning (ML) models can develop a deep understanding of a given domain, leading to breakthroughs for top tech companies. Google, for instance, fine-tunes its ranking algorithms by tracking and analyzing more than one trillion search queries each year. It turns out that the Solomonic power to answer all questions from all comers can be brute-forced with sufficient data.

But there’s a catch: Most companies are limited to “small” data; in many cases, they possess only a few dozen examples of the processes they want to automate using ML. If you’re trying to build a robust ML system for enterprise customers, you have to develop new techniques to overcome that dearth of data.

Two techniques in particular — transfer learning and collective learning — have proven critical in transforming small data into big data, allowing average-sized companies to benefit from ML use cases that were once reserved only for Big Tech. And because just 15% of companies have deployed AI or ML already, there is a massive opportunity for these techniques to transform the business world.

2020-10-08 00:00:00 Read the full story…
Weighted Interest Score: 2.6248, Raw Interest Score: 1.8855,
Positive Sentiment: 0.2076, Negative Sentiment 0.1730

5 Concepts Every Data Scientist Should Know

Once a Data Scientist, there are certain skills you will apply each and every day of your career. Some of these might be common techniques you learned during your education, while others may develop fully only after you become more established in your organization. Continuing to hone these skills will provide you with valuable professional benefits.

I have written about common skills that Data Scientists can expect to use in their professional careers, so now I want to highlight some key concepts of Data Science that can be beneficial to know and later employ. I may be discussing some that you know already and some that you do not know; my goal is to provide some professional explanation of why these concepts are beneficial regardless of what you do know now. Multicollinearity, one-hot encoding, undersampling and oversampling, error metrics, and lastly, storytelling, are the key concepts I think of first when thinking of a professional Data Scientist in their day-to-day. The last point, perhaps, is a combination of skill and a concept but wanted to highlight, still, its importance on your everyday work life as a Data Scientist. I will expound upon all of these concepts below.

  1. Multicollinearity
  2. One-Hot Encoding
  3. Sampling
  4. Error Metrics
  5. Storytelling

2020-10-05 00:00:00 Read the full story…
Weighted Interest Score: 2.4810, Raw Interest Score: 1.2642,
Positive Sentiment: 0.2054, Negative Sentiment 0.2845


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 customer_success@cloudquant.com. 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.