AdobeStock_91734599

AI & Machine Learning News. 15, October 2018

NHS hospital is developing AI cameras that can spot intruders and monitor staff

Great Ormond Street Hospital (GOSH) has paired up with Cambridge-based technology giant Arm to develop smart “person recognition” cameras that can spot intruders and monitor staff on its premises, in a move that could speed up operations and improve security.

The artificial intelligence technology will be used by the children’s hospital in London to track doctors and nurses on site, whilst also keeping a watch on visitors as they pass from the reception area to wards.
2018-10-10 00:00:00 Read the full story.

CloudQuant Thoughts… GOSH is a great hospital and like others around the world it struggles with keeping an open friendly environment whilst warding off intruders. Unlike many of the examples we have seen so far, this is an example of AI based human movement tracking to which no-one could object.

 

Finally, A Disruptor to Uber…Maybe

Google’s Waze has been making waves by using its data about cars on the road to assemble the largest carpooling platform ever. This has the potential to disrupt Uber. On Waze, you can schedule a ride with a driver going the same direction as you. You can see reviews of riders, drivers, and make a new friend, all while keeping a car off the road. Environmentally, it is smart. Economically, it leverages slack resources at virtually no cost. Drivers get “some gas money,” so that is even less than hailing an Uber. Frankly, it is a brilliant way to use data to share in the activity of driving and commuting. If this catches on, it has some big implications:

  1. Winning Technology Always Lowers Price
  2. Long-term Car Demand is Uncertain and Declining
  3. Making Boring Things Fun is…Well, Fun
  4. This is Bad for Mass Transit
  5. The HOV Lanes are Going to Get Full

I just got a Waze driver to take me to work for less than $2.00, which is about 10% of the Uber price. This is great! Bye-bye, Uber…well maybe.
2018-10-11 00:00:00 Read the full story.
CloudQuant Thoughts… I used Waze a lot when it first came out but drifted away from it. I will have to take another look. This is major distruption in play!

 

M.I.T. Plans College for Artificial Intelligence, Backed by $1 Billion

Every major university is wrestling with how to adapt to the technology wave of artificial intelligence — how to prepare students not only to harness the powerful tools of A.I., but also to thoughtfully weigh its ethical and social implications. A.I. courses, conferences and joint majors have proliferated in the last few years. But the Massachusetts Institute of Technology is taking a particularly ambitious step, creating a new college backed by a planned investment of $1 billion. Two-thirds of the funds have already been raised, M.I.T. said, in announcing the initiative on Monday. The goal of the college, said L. Rafael Reif, the president of M.I.T., is to “educate the bilinguals of the future.” He defines bilinguals as people in fields like biology, chemistry, politics, history and linguistics who are also skilled in the techniques of modern computing that can be applied to them.
2018-10-15 00:00:00 Read the full story.
CloudQuant Thoughts… Students of computer science are notoriously single minded in their academic pursuit, Students of other disciplines likewise. This College opens up opportunities for “wide thinkers”, students and teachers who may not be rigorous enough for either individual line of study but will likely be the ultimate creators of our future.

 

Corti heart attack detection AI can now deploy on the edge with Scandinavian design

Work is underway to deploy Corti, an AI system that detects heart attacks during emergency phone calls, and it could be coming to some of the biggest cities in Europe.

…this summer the European Emergency Number Association (EENA) will deliver AI-powered assistance to emergency 112 operators. In initial trials, this assistance was found to identify cardiac arrest events more quickly than human operators.

Emergency call centers from Seattle to Singapore also want to make Corti part of their operations, but there’s no global standard for organizations working to save lives. Some are fine with the idea of deploying the AI through the cloud, while others with privacy concerns require the AI system to operate from on-premise servers.
2018-10-14 00:00:00 Read the full story.
CloudQuant Thoughts… Week after week we see stories of how AI is helping to bring new humans into the world and save humans from an early death. Articles like this show AI’s full benefit to the human race.

 

Google Cloud drops out of the running for the Pentagon’s $10B cloud contract

Google’s decision to articulate a set of principles around its use of artificial intelligence has officially cost it business. The company announced Monday that it will not submit a bid for the U.S. Department of Defense’s JEDI cloud computing project, a massive undertaking to transform the military’s information technology infrastructure that is expected to be worth at least $10 billion over the next ten years. A Google representative confirmed a Bloomberg report that it will not bid on the contract, months after employee unrest over its involvement with Project Maven — an attempt to use artificial intelligence to better identify the targets of drone strikes — led to its decision to let the contract covering that work expire.
2018-10-08 21:54:39-07:00 Read the full story.

Microsoft workers protest bid to build Pentagon’s $10bn AI warfare system

Microsoft employees have signed a letter calling on executives to ditch a proposed bid for a US military contract that would see the technology giant providing computer power for artificial intelligence to be used in warfare for the next ten years. The letter accuses of Microsoft of betraying its principles “in exchange for short-term profits” in a plan that would force employees to build a product who have no idea “whether our work is being used to aid profiling, surveillance, or killing”.

“Many Microsoft employees don’t believe that what we build should be used for waging war. When we decided to work at Microsoft, we were doing so in the hopes of ’empowering every person on the planet to achieve more,’ not with the intent of ending lives and enhancing lethality,” the letter continued.

Little is known about what the government plan to use the proposed cloud system, named the Joint Enterprise Defense Infrastructure (JEDI). At an industry conference to discuss JEDI, the US Defence Department’s chief management officer John Gibson revealed that the controversial program was “truly about increasing the lethality of our department”.
2018-10-13 00:00:00 Read the full story.
CloudQuant Thoughts… The battle hots up between the bottom-line and the will of the employees. How it will play out nobody knows, except maybe SkyNet.

 

End of the Line for Buy-Side OMS?

Buy-side order management systems may have reached an evolutionary dead end as trading automation and multi-asset trading continues to grow.

“The OMS has evolved into a sort of all-in-one system that, by its nature, was bound to hit limitations in terms of servicing every need of a portfolio manager, the middle office, operations, compliance and the buy-side trade,” John Adam, senior director, portfolio management & trading solutions at FactSet, told Markets Media.

In a recent white paper written by consultancy Alignment Systems and published by FactSet, the author divided OMS platforms into one of five generations with each generation growing in sophistication :

  • Generation Zero : routed orders to asset-class specific desks
  • Generation One : supports simple rules that would incorporate the use of static data
  • Generation Two : uses more advanced rules integrating real-time data
  • Generation Three : supports more complex rules and would use historical analysis
  • Generation Four : supports machine learning.

2018-10-08 20:44:31+00:00 Read the full story.
CloudQuant Thoughts… Buy side OMS vendors that are innovating into making fund management easier with integration to execution, allocations (pre-trade and post-trade) will continue for the foreseeable future. The buy side still needs these features. As hedge funds become more sophisticated, with data science tools like JupyterLab and Python, the innovative vendors will provide new empowering capabilities. We don’t think this is the “end of the line,” rather we are convinced it is a springboard.

 

 

The Top 3 NIO Shareholders

NIO Inc. (NIO), founded in 2014 by Chinese entrepreneur William Li, is a Shanghai-based automobile manufacturer that specializes in designing and developing electric autonomous vehicles. The company, formerly known as NextEV, has big plans to become the Tesla Inc. (TSLA) of China and satisfy Beijing’s ambitions to rapidly expand the country’s production of sophisticated technologies, including vehicles that reduce emissions.

In September, NIO became the third-biggest U.S. listing by a Chinese firm this year. The company raised $1 billion in its initial public offering (IPO), falling short of initial expectations. The shares were priced at $6.26 a piece on Sept. 12, just above the low end of its $6.25 to $8.25 target price range. Sources told Reuters that its valuation was dragged down by investor concerns about the troubles engulfing chief competitor Tesla. Waning sentiment in Chinese companies probably also did not help.

At the start of the IPO process, NIO had hoped for a valuation of as much as $20 billion. In the end it had to settle for a market capitalization of $6.41 billion. The shares currently trade at $7.39, with a market cap of $7.52 billion.

Top 3 shareholders : William Li, Tencent, Baillie Gifford (the second-largest holder of Tesla stock after Elon Musk).
2018-10-11 01:58:00-06:00 Read the full story.

CloudQuant Thoughts… There was a lot of energy generated by this the release of this information last week. Was Baillie Gifford diversifying, shifting or covering all possible outcomes? The release caused a $1 move, not bad for a $6 stock. Can you scrape websites for “shareholder movement” and use that to predict stock price action? Perhaps we already have data that can give you a steer in the right direction in one of our Alternative Data Sets. Check out CloudQuant.com and show us what you can do.

 


RAPIDS – The news of the week….

NVIDIA Introduces RAPIDS Open-Source GPU-Acceleration Platform for Large-Scale Data Analytics and Machine Learning

NVIDIA today announced a GPU-acceleration platform for data science and machine learning, with broad adoption from industry leaders, that enables even the largest companies to analyze massive amounts of data and make accurate business predictions at unprecedented speed.

RAPIDS™ open-source software gives data scientists a giant performance boost as they address highly complex business challenges, such as predicting credit card fraud, forecasting retail inventory and understanding customer buying behavior. Reflecting the growing consensus about the GPU’s importance in data analytics, an array of companies is supporting RAPIDS — from pioneers in the open-source community, such as Databricks and Anaconda, to tech leaders like Hewlett Packard Enterprise, IBM and Oracle.

2018-10-09 00:00:00 Read the full story.

 

Nvidia Platform Pushes GPUs into Machine Learning, High Performance Data Analytics

With Nvidia’s effort to push the GPU acceleration into ML/HPDA (high performance data analytics), the company reports that the RAPIDS platform delivers speed-ups, using the XGBoost machine learning algorithm for training on an NVIDIA DGX-2 supercomputer, of 50x compared with CPU-only systems.

RAPIDS brings with it with an ecosystem from the open-source community, including Databricks (a web-based platform for big data processing in the cloud using Apache Spark) and Anaconda (an open source distribution of the Python and R programming languages for data science and machine learning), and tech companies such as Hewlett Packard Enterprise, IBM and Oracle.
2018-10-10 00:00:00 Read the full story.

 

Bringing Dataframe Acceleration to the GPU with RAPIDS Open-Source Software from NVIDIA

RAPIDS is the culmination of 18 months of open source development to address a common need in data science: fast, scalable processing of tabular data for extract-transform-load (ETL) operations. ETL tasks typically are not GPU accelerated, but recent developments in both hardware and software brought together by RAPIDS make it practical for data scientists to accelerate ETL on NVIDIA GPUs. Data wrangling is the new bottleneck : The explosive growth of GPU performance on model training tasks has made ETL stages in the data science workflow the new bottleneck. We’ve seen an explosion of GPU-accelerated deep learning use cases, like image recognition, document classification, and time series prediction. But any data scientist will tell you a model is only as good as the data you put into it. Preparing training and validation data is often as much work as designing and training the model itself. Loading, selecting, transforming, and aggregating data is an important part of the data science workflow, and seldom done only once. When these ETL tasks are combined during data exploration, feature engineering, and model verification, the data scientist often will need to do these tasks over and over again with increasingly larger datasets, at which point performance becomes important.

GPU-accelerated Pandas…
2018-10-10 10:00:57+00:00 Read the full story.

 

IBM and NVIDIA Collaborate to Expand Machine Learning Tools for Data Scientists

A new press release reports, “IBM today announced that it plans to incorporate the new RAPIDS™ open source software into its enterprise-grade data science platform for on-premises, hybrid, and multicloud environments. With IBM’s vast portfolio of deep learning and machine learning solutions, it is best positioned to bring this open-source technology to data scientists regardless of their preferred deployment model. ‘IBM has a long collaboration with NVIDIA that has shown demonstrable performance increases leveraging IBM technology, like the IBM POWER9 processor, in combination with NVIDIA GPUs,’ said Bob Picciano, Senior Vice President of IBM Cognitive Systems. ‘We look to continue to aggressively push the performance boundaries of AI for our clients as we bring RAPIDS into the IBM portfolio’.” The release goes on, “RAPIDS will help bring GPU acceleration capabilities to IBM offerings that take advantage of open source machine learning software including Apache Arrow, Pandas and scikit-learn. Immediate, wide ecosystem support for RAPIDS comes from key open-source contributors.”
2018-10-11 00:05:21-07:00 Read the full story.

 

Oracle and NVIDIA Partner on Analytics, Machine Learning, and AI

Oracle and NVIDIA have announced that Oracle is the first public cloud provider to support the NVIDIA HGX-2 platform on Oracle Cloud Infrastructure , helping to meet the needs of the next generation of analytics, machine learning and artificial intelligence (AI). The companies are also announcing the general availability of support for GPU-accelerated deep learning and high performance computing (HPC) containers from the (NGC) container registry …
2018-10-10 00:00:00 Read the full story.

 


Below the Fold…

What Can Multiparadigm Data Science Do For You? Wolfram Research (Video 5 mins)

Many organizations are still doing traditional data science — confining themselves to problems that are answerable with traditional statistical methods — rather than utilizing the broad range of interfaces and techniques available today. Multiparadigm Data Science includes automated machine learning, interactive notebooks and report generation, natural language queries of data for instant visualizations, and implementing neural networks with ease and efficiency. – Erez Kaminski, Wolfram Technology Specialist, Global Technical Services, Wolfram Research, Inc., at Data Summit 2018.
2018-10-09 00:00:00 Read the full story.

 

To solve more problems AI needs to think small

Many small and medium businesses don’t have the luxury of big data. They need the tools to generate rational and informed insight from small data sets. Unless these small volumes can be exploited there is a risk that AI will fail to deliver on its promise. AI itself is much broader than machine learning and was making an impact way before the data surge. It would be remiss to forget the useful contributions and tools available that can solve problems where data is scarce. “Working with little data has been the reality of AI for over 40 years. There are lots of methods, like SVMs, decision trees, regressions, probabilistic models, and others, that work very well on “small” datasets,” he explains.

Big data is regularly criticised for its overriding faith in ‘correlation is king’. Algorithms alone lack the causal inference and extrapolation capabilities to deliver informed and rational solutions from big data sets, although often throwing more data at a problem works in the end. This poses a problem for big data’s pint-sized variants. Surely relying on small data will just attenuate meaningful insight? If these somewhat reckless approaches are applied to all problems, we may end up being disappointed by AI in most cases. Garcia-Gasulla disagrees. He argues AI is capable of generating rational and informed insight from small data volumes if researchers bring outmoded data-analysis techniques back to the fore.
2018-10-11 00:00:00 Read the full story.

 

Harnessing Artificial Intelligence to Create New Financial Products – (Video 4 mins)

The World Economic Forum’s Jesse McWaters discusses how large financial firms are using artificial intelligence to address gaps in their product offerings. Even mid-size firms, such as China’s Ping An, are using the technology to sift through troves of data and create new products.
2018-10-12 16:47:03-04:00 Read the full story.

 

Microsoft Extends ML Framework

The latest release of Microsoft’s machine learning framework incorporates a new API intended to help .NET developers train models for improved predictions in uses cases like image classification and speech translation. Microsoft said the 0.6 version of its ML.NET machine learning framework released this week expands the number of data pipelines that can be used to build machine learning models. Previous versions limited the types of pipelines that could be used to train models. The new version also improves model prediction performance, the company said this week. An earlier version of ML.NET added support for TensorFlow models with the goal of using deep learning models to improve prediction performance for uses cases such as image classification, speech to text and translations. The new version adds support for predictions derived from the Open Neural Network Exchange format, which is billed as an open platform for interchangeable AI models.
2018-10-10 00:00:00 Read the full story.

 

Using Data Lakes to Fuel Self-Service Analytics

Data lakes have helped organizations deal with the massive amounts of data generated daily. They are intended to serve as a central repository for raw data, a treasure trove for data scientists to analyze and gain actionable insight. They also serve as the foundation for many “self-service” analytics initiatives. The promise of a data lake includes a pristine environment that is easy to use and gives organizations the opportunity to use machine learning to mine the true value of their data. The insight gleaned will help increase competitiveness and improve decision making. One of the main attractions of data lakes is flexibility. Getting data into a lake is simple. Getting insight and value from all of that data, however, has proven to be challenging. A recent Forrester report found that 60%–73% of all enterprise data goes unused for analytics. This statistic exposes some of the harsh realities of data lakes.

The biggest hurdle with a data lake takes place during its inception. It’s quite easy, and common, to misalign on the purpose and content of the lake. In users’ eagerness to aggregate data, they have often overlooked establishing the processes and controls that address key questions about the data: What is it? Who owns it? How should it be defined? Does it really belong in the lake? Instead, every bit and byte of data has found its way into the lake, and this lack of oversight has muddied the waters. As a result, it’s difficult for users to know what’s in the lake, or if the data they do find can be trusted.
2018-10-10 00:00:00 Read the full story.

 

Governance of ML Models in a Bank

Machine learning (ML) adds a unique and sometimes overwhelming experience to a decision system. The use cases that fit for using Machine Learning approach in a bank are numerous. Usually, in banks, the departments have their own IT teams to carry out enhancements & maintenance of existing systems. Even in those banks which have adopted the shared services model for IT teams have segregated the resources based on the department’s funding and end up segregating the Shared services IT team into SME groups.

So, a bank if now looks at AI as the next frontier of the problem solving, then it brings a very interesting challenge to the picture… Governance.
2018-10-13 12:55:25 Read the full story.

 

DBA CORNER – The Mainframe Does Big Data

…most organizations are planning to use mainstream relational databases to support their big data environment—more so than Hadoop, NoSQL, or any other type of database or data platform. So traditional RDBMSs, such as Db2 for z/OS, can be—and are being—used to drive big data projects. O’Reilly’s Data Science Salary Survey found that the top tool used by data scientists is not R or Python, but SQL on relational databases. Yes, the same SQL we all know (and love) has not been displaced by other tools for big data analytics. It makes sense that relational databases can be used for big data projects, but probably not all big data projects. Sometimes relational may not operate or perform as needed when the project requires a large number of columns, a flexible schema, relaxed consistency, or complex graphs. But that does not mean the mainframe cannot be used.
2018-10-10 00:00:00 Read the full story.

 

Knowledge Plus Statistics: Understanding the Emerging World of Deep Probabilistic Programming Languages (Deep PPLs)

The use of statistics to overcome uncertainty is one of the pillars of a large segment of the machine learning market. Probabilistic reasoning has long been considered one of the foundations of inference algorithms and is represented is all major machine learning frameworks and platforms. Recently, probabilistic reasoning has seen major adoption within tech giants like Uber, Facebook or Microsoft helping to push the research and technological agenda in the space. Specifically, probabilistic programming languages(PPLs) have become one of the most active areas of development in machine learning sparking the release of some new and exciting technologies. Conceptually, probabilistic programming languages(PPLs) are domain-specific languages that describe probabilistic models and the mechanics to perform inference in those models. The magic of PPL relies on combining the inference capabilities of probabilistic methods with the representational power of programming languages.

In a PPL program, assumptions are encoded with prior distributions over the variables of the model. During execution, a PPL program will launch an inference procedure to automatically compute the posterior distributions of the parameters of the model based on observed data. In other words, inference adjusts the prior distribution using the observed data to give a more precise mode. The output of a PPL program is a probability distribution, which allows the programmer to explicitly visualize and manipulate the uncertainty associated with a result.
2018-10-15 12:59:25.905000+00:00 Read the full story.

 

Chinese tech giant Huawei unveils A.I. chips, taking aim at giants like Qualcomm and Nvidia

Huawei unveiled two new artificial intelligence chips aimed at data centers and smart devices, pitting it against major silicon players including Qualcomm and Nvidia, as the Chinese giant laid out a strategy it hopes will drive growth in the next few years. The new chipsets are called the Ascend 910 and Ascend 310 and were revealed Wednesday at the Huawei Connect conference in Shanghai, China.

Huawei’s Ascend 910 is aimed at data centers. Companies using AI applications require huge amounts of data to train smart algorithms, which can take several days or weeks. Huawei claims that its chip can process more data in a faster amount of time than its competitors and help train networks in a matter of minutes.
2018-10-10 00:00:00 Read the full story.

 

AI delivers a paradigm shift for credit management

Embracing and implementing AI-based technologies can make functions such as credit management more efficient and cost-effective, while minimizing risk along the way, writes Sébastien Méric, chief innovation officer at Tinubu Square. “We’ve always done it this way…” A common enough refrain in many industries, but if we allowed it to set the pace, no company would ever get out of its comfort zone, never evolve, and certainly never improve.

Artificial intelligence (AI) is just one example of innovation that has a massive part to play. Put simply AI involves computer programs that are able to automatically generate cognitive functions. In practical applications, it specifically involves robots that can adapt to situations not initially predicted by the engineers who designed them. Therefore, a robot must be able to adapt its behavior within situations in which it has never before operated, relying on experience or knowledge from past experiences.
2018-10-12 00:00:00 Read the full story.

 

3 Ways AI In The Business World Can Lead To Industry Improvement

Artificial intelligence has disrupted countless industries, but adapting AI in the business world is a slow process. Here’s how it stands to change things.

  1. Virtual assistance
  2. Generating insights
  3. Automation of the manual process

2018-10-10 21:14:17+00:00 Read the full story.

 

Better Buy: International Business Machines Corporation vs. McDonald’s — The Motley Fool

At first glance it may seem odd to compare International Business Machines (NYSE:IBM) and McDonald’s (NYSE:MCD), as one company is a legacy tech giant and the other is the world’s largest fast-food chain. Beyond the obvious differences, however, the two companies have a number of similarities as stocks. They are both blue-chip members of the Dow Jones Industrial Average, and they’ve both been around for generations. McDonald’s dates back to 1940, while IBM was founded in 1911. The two stocks also check many of the same boxes for investors, as they are both reliable dividend payers — though only McDonald’s is a Dividend Aristocrat, meaning it’s raised its dividend every year for 25 years in a row. IBM, however, is only a few years away from achieving that status. As stocks, both are down modestly this year, though McDonald’s has been the clear winner over the last five years, as the chart below shows.
2018-10-15 00:00:00 Read the full story.

 

Ultimate Privacy Will Soon Be Lost

Companies often grapple with the challenges of respecting customer privacy. Although most firms do not seek to cause or do harm, data breaches, inadvertent misuse of data, and unauthorized publishing of data seem to ultimately happen. Data on a social media site, on a shopping site, or from your gym membership may seem sensitive. But, the most sensitive data is surely one’s own, unique DNA. Now, DNA is being publishing, shared, and studied in great ways. Over 1 million Americans have shared their DNA with genealogical sites and many hundreds more do so each week. Researchers published in the Journal of Science have calculated that over 60% of the people with European descent can be linked to a third cousin or closer. It is a staggering thought and reminds us that we ultimately share a common heritage and DNA explains all of our difference and similarities.

Recently, law enforcers used DNA submitted by family members to genealogical sites to identify and charge the Golden State killer. Another 13 murder suspects have been found simply through clever DNA sloughing. This is a great advance for human society – no crime is without reach. DNA long family searching will soon be a standard procedure for investigating crimes.

However, what are the implications for fraud, privacy, and even the control of our own precious, unique DNA? With the DNA available online and a few pieces of information like possible age and place or birth, we can all be found through our family trees. That is amazing. Can this be used or abused by insurers? Perhaps it can be used to find better organ donations? Still, the challenge is that we have lost control or never seceded over control of this important personal marker to groups that may indeed use it without our best interest in mind.
2018-10-13 00:00:00 Read the full story.

 

Kubernetes Is a Prime Catalyst in AI and Big Data’s Evolution

Kubernetes is becoming synonymous with cloud-native computing. As an open-source platform, it enables development, deployment, orchestration and management of containerized microservices across multicloud ecosystems. Kubernetes is the key to cloud-native microservices that are platform agnostic, dynamically managed, loosely coupled, distributed, isolated, efficient, and scalable. The maturation of Kubernetes continues to deepen as it leverages containers, orchestrations, service meshes, immutable infrastructure, and declarative APIs. One clear indicator of Kubernetes’ maturation is the rich ecosystem of other open-source projects that have grown up around it…
2018-10-10 00:00:00 Read the full story.

 

NEXT-GEN DATA MANAGEMENT – Dangers of Statistical Modeling

As we enter a world of machine learning and data science, are there any gotchas or negatives? It sounds as if it is all sunshine and rainbows, but, as the title to this post alludes, I believe there are. Here are some of the dangers I thought of or came across while researching this post:

  • Normal does not equal good.
  • Beware statistical bias.
  • We stop thinking and just believe the model is right.
  • Statistical models can suffer from the “boiling frog” syndrome.

Let’s take a closer look and see how each one could ultimately work against us.
2018-10-10 00:00:00 Read the full story.

 

PodCast : DataHack Radio #12: Exploring the Nuts and Bolts of Natural Language Processing with Sebastian Ruder

There’s text everywhere around us, from digital sources like social media to physical objects like books and print media. The amount of text data being generated every day is mind boggling and yet we’re not even close to harnessing the full power of natural language processing. I see a ton of aspiring data scientists interested in this field, but they often turn away daunted by the challenges NLP presents. It’s such a niche line of work, and we at Analytics Vidhya would love to see more of our community actively participate in ground-breaking work in this field.

So we thought what better way to do this than get an NLP expert on our DataHack Radio podcast? Yes, we’ve got none other than Sebastian Ruder in Episode 12! This podcast is a knowledge goldmine for NLP enthusiasts, so make sure you tune in. It’s catapulted to the top of my machine learning recommended podcasts. I strongly feel every aspiring and established data science professional should take the time to hear Sebastian talk about the diverse and often complex NLP domain.
2018-10-14 23:17:21+05:30 Read the full story.

 

Amazon Recruiting Snafu Shows Dangers of Machine Learning

Like many tech companies, Amazon has experimented with machine learning (ML) techniques to improve its recruiting process. But according to a new Reuters report, the company hit a huge snag a few years ago: Its algorithms began preferring male applicants. Reuters based its report on five anonymous sources, who said that Amazon’s ML-based tool rated candidates on a scale of one to five stars. The tool’s algorithms analyzed 10 years’ worth of résumés and assumed that, because the majority of applicants were male, males were preferable hires. As a result, it began downgrading any mention of “women” or “women’s” in résumés. “Amazon edited the programs to make them neutral to these particular terms,” Reuters added. “But that was no guarantee that the machines would not devise other ways of sorting candidates that could prove discriminatory, the people said.” Eventually, Amazon decided to dissolve the team behind the project.
2018-10-11 00:00:00 Read the full story.

Amazon ditches AI recruitment tool that ‘learnt to be sexist’

Amazon has scrapped a “sexist” internal tool that used artificial intelligence to sort through job applications. The program was created by a team at Amazon’s Edinburgh office in 2014 as a way to sort through CVs and pick out the most promising candidates. However, it taught itself to prefer male candidates over female ones, members of the team told Reuters.

They noticed that it was penalising CVs that included the word “women’s”, such as “women’s chess club captain”. It also reportedly downgraded graduates of two all-women’s colleges. The problem stemmed from the fact that the system was trained on data submitted by people over a 10-year period, most of which came from men.
2018-10-10 00:00:00 Read the full story.

 

Deep Learning Performance Cheat Sheet – Towards Data Science

The question that I get the most from new and experienced machine learning engineers is “how can I get higher accuracy?” Makes a lot of sense since the most valuable part of machine learning for business is often its predictive capabilities. Improving the accuracy of prediction is an easy way to squeeze more value from existing systems. The guide will be broken up into four different sections with some strategies in each:

  • Data Optimization
  • Algorithm tuning
  • Hyper-Parameter Optimization
  • Ensembles, Ensembles, Ensembles

2018-10-15 12:55:20.337000+00:00 Read the full story.

 

Defending data centres requires an Artificial Intervention

We don’t talk about it. If mentioned it’s a little over a whisper. The dirty little phrase that no CIO wants to ever hear but everybody knows is here to stay: ‘We’ve just been hacked’. The threat and cost to businesses are very real. How real, first came to my attention when in 2015 Talk Talk announced they had been hacked with 100,000 customers going into a panic and leaving, costing the company over £60 million. That loss is minuscule compared to what Equifax suffered thanks to their security breach last year – a whopping $4 billion according to Time Magazine. We are still yet to know how much the latest cyber security breach will cost British Airways.

The way data enters, computes, is stored, and then leaves the data centre has changed immeasurably over the years. The 2018 Cyber Security Breaches Survey found that in the UK, 43% of businesses had reported cyber security breaches or attacks in the last 12 months rising to 72% among large businesses.
2018-10-11 00:00:00 Read the full story.

 

Tencent executive urges Europe to focus on ethical uses of artificial intelligence

Chinese tech giant Tencent has urged European companies to focus on ethical applications of artificial intelligence, leaving higher-risk ventures to the US and China. Speaking at a conference in Helsinki, Finland, David Wallerstein, Tencent’s chief exploration officer, said he was encouraging the European Union to “embrace AI and deploy it in the areas that would have a maximum benefit for human life, even if that technology isn’t competitive to take on an American or Chinese market”.

“By the time you get better at AI in Europe, the planet will have 8.5 billion people and most of the additional billion will be in the developing world. Energy is an area where there’s a huge opportunity on the planet, and it’s a huge opportunity for Europe. “I’ve heard lots of people saying how do we catch up with China and the US in the next 15 years, but we may not have much of a planet left by then,” he said.
2018-10-14 00:00:00 Read the full story.

 

Brain-Inspired Cognitive Architecture Is Now Solving Computational Challenges Faced By AI

With the development of artificial intelligence intensifying across the globe, IT companies are looking for ways to revamp their architecture to make more robust. Increasingly, researchers are turning to brain-inspired architecture with co-located memory and processing, resulting in computers which are 200 times faster than conventional computers. Such is the excitement around AI hardware, that this phase has been dubbed as a “renaissance of hardware” as vendors are rushing to build domain-specific or workload-specific architectures that can significantly scale and improve computational efficiency.

And as we nudge forward in the mobile era, the workloads are going to look extremely dissimilar since the requirements of computing are changing. Businesses have to rely on a different architecture, each meant for a particular workload. This is where vendors are making a shift from Von Neumann computing architecture and are striving to improve the performance of computing with multi-core CPU architectures.

Now, the rapid gains in neuroscience have also spurred researchers across the globe to propose Brain-inspired computing architecture to develop highly advanced cognitive systems. IBM researchers are working on a new computer architecture which can process data efficiently for AI workloads. However, what’s remarkable is that this new architecture is inspired by the brain and will feature coexisting processing and memory units.
2018-10-12 12:35:52+00:00 Read the full story.

 

Artificial Intelligence & Marketing: Best Friends Yet?

…what are the biggest challenges that marketers face today? Are any of these problems solvable at scale through AI? Few of the problems that have plagued marketers─those who that don’t have gargantuan budgets and resources to solve these perpetual problems are outlined below:

  1. Knowing the customer DNA: 360-view of the customer
  2. Understanding intent in search: The evolution of recommendation engines
  3. Incorrect forecasting: How AI can help improve and align marketing forecasting models
  4. Voice Commerce: How can AI help?
  5. Privacy laws & data protection post GDPR

2018-10-12 12:51:32+00:00 Read the full story.

 

Student Project for Monitoring of Movement Disabilities with Movesense and Azure IOT Hub

Sophia Botz, Computer Science student at University College London, has been working with Microsoft and Great Ormond Street Hospital in London on her IXN Project. Sophia has developed a health related Xamarin app which use Movesense sensors, with the Xamarin plugin for Movesense and Azure IoT Hub infrastructure to be analysed by an Azure Machine Learning service.

The project was proposed by Great Ormond Street Hospital in collaboration with Microsoft for her MSc Computer Science at UCL. The aim of the project was to develop a proof of concept for a system that can record motion data and classify this into different movements in order to be used in the monitoring and treatment of movement disabilities.
2018-10-15 00:00:00 Read the full story.

 

Weekly Selection — Oct 12, 2018 – Towards Data Science

 

  • The 4 Convolutional Neural Network Models That Can Classify Your Fashion Images
  • How to Data Science without a Degree
  • Large-scale Graph Mining with Spark
  • How Taxis Arrive at Fares? — Predicting New York City Yellow Cab Fares
  • How to make a gif map using Python, Geopandas and Matplotlib
  • An Intuitive Explanation of Policy Gradient — Part 1: REINFORCE
  • Your Ultimate Guide to Matplotlib
  • Collaborative Embeddings for Lipstick Recommendations
  • Stop Installing Tensorflow using pip for performance sake!

2018-10-12 Read the full story.

 

The 4 Convolutional Neural Network Models That Can Classify Your Fashion Images

Clothes shopping is a taxing experience. My eyes get bombarded with too much information. Sales, coupons, colors, toddlers, flashing lights, and crowded aisles are just a few examples of all the signals forwarded to my visual cortex, whether or not I actively try to pay attention. The visual system absorbs an abundance of information. Should I go for that H&M khaki pants? Is that a Nike tank top? What color are those Adidas sneakers?

Can a computer automatically detect pictures of shirts, pants, dresses, and sneakers? It turns out that accurately classifying images of fashion items is surprisingly straight-forward to do, given quality training data to start from. In this tutorial, we’ll walk through building a machine learning model for recognizing images of fashion objects using the Fashion-MNIST dataset. We’ll walk through how to train a model, design the input and output for category classifications, and finally display the accuracy results for each model.

2018-10-06 Read the full story.

 

How to Data Science without a Degree

I want to show you how to become a Data Scientist without a degree (or for free). Ironically, I do have a degree — one that was even made for Data Science (Master’s in Analytics from Northwestern). But to give you a little background, I used to be an accountant at Deloitte. Isn’t that crazy? I was far from data science or anything technical. I had to learn a lot of things online on my own after work and even during my Master’s program to catch up to my peers’ level as I came from a non-technical background. Having gone through the experience myself, I can tell you that degree is very helpful, but not necessary. Because I have been on both sides of getting a degree and learning things online, I think I can give you a unique perspective. Getting a Master’s in data science is a sure and fast way to get into the field, but luckily you don’t have to if you don’t want to spend $60–90k on tuition. It will require a lot of your self-discipline though.

2018-10-05 Read the full story.

 

How to make a gif map using Python, Geopandas and Matplotlib

As a language, Python is enormously flexible. Which makes it possible to make lots of different visualisations in, at times, only a few lines of code. But with all the different charting websites and software available now, why bother writing code? Can’t we just use a GUI to upload a csv file, adjust the range, hit export png and call it a day?

Yes. You certainly can. And sometimes, this is the best choice if you need a quick, one-time chart or map. But the real power of using Python comes into play when you need to make lots of maps — lots and lots of maps.

2018-10-10 Read the full story.

 

For SAP CX, Artificial Intelligence Starts with Personalisation

The move to build AI capabilities into application software has been one of the most important developments to wash through the technology sector in the last few years. Whether it is Adobe’s Marketing Cloud and Sensei, or Salesforce’s CRM and Einstein, the integration of AI into business functionality has arrived surprisingly quickly – after decades of incubation in universities and computer science labs.

SAP’s Chris Hauca, Head of Strategy and GTM, SAP Enterprise Commerce, SAP Customer Experience, spoke to Which-50 at the recent SAP Customer Experience Live conference in Barcelona and we started by asking where his company was putting its initial efforts.
2018-10-11 23:22:37+11:00 Read the full story.

 

This data analytics team uncovered $3.6B in healthcare savings in a year

There is no shortage of opportunity for data analytics leaders in healthtech — especially one with as much experience as Matthew Beatty, who built a 20-year career in the industry prior to joining Payformance Solutions as director of analytics in late 2017. Payformance’s TrustHub platform was developed to make it easier for healthcare payers and providers to transition from fee-for-service payment models to value-based ones. Joining the startup gave Beatty the chance to tackle a massive technical challenge — and make a real impact.

“Changing the way healthcare is paid for drives changes on how healthcare is delivered,” said Beatty. “Healthcare analytics is not just about lowering costs. It’s about bending the cost curve and using the savings in transformative ways, such as buying air conditioners for low-income patients with asthma or launching nutrition programs for diabetics.”
2018-10-10 00:00:00 Read the full story.

 

One To One — Digital Marketing Conference In Biarritz

Thomas Husson, Vice President, Principal Analyst, Forrester – I am very excited to have been invited for the third time to give the opening keynote speech at One to One, a great marketing conference taking place on October 10–12 in Biarritz in the southwest of France.

  • AI-powered conversations do not really exist yet.
  • Beyond the tech hype, CMOs should embrace conversational marketing.
  • AI will spark a marketing renaissance.
  • Marketers in France struggle to deliver a good customer experience.

2018-10-10 15:51:31-04:00 Read the full story.

 

Amazon hires top UW computer science prof as new robotics director, a year after he arrived from CMU

Amazon is tapping one of the world’s leading human-robot interaction experts to help manage a robotics operation that has become an integral part of the company’s fulfillment centers. The Seattle tech giant has hired Siddhartha “Sidd” Srinivasa, a computer science professor at the University of Washington, as its director of robotics.

Srinivasa arrived in Seattle last year after moving his entire team of more than a dozen researchers from Carnegie Mellon University to the UW. He became the latest renowned technologist to join the top-ranked University of Washington computer science program, which will house Srinivasa’s team in its new building slated to open in January. Srinivasa will split time between his new role at Amazon and the University of Washington. At Amazon, Srinivasa will work with Brad Porter, currently vice president and distinguished engineer of robotics.
2018-10-12 17:47:41-07:00 Read the full story.

 

MPs invite robot to give evidence on AI

A robot is set to become the first non-human to appear as a witness before the UK Parliament. The Commons Education Select Committee invited Pepper the robot from Middlesex University to give evidence at a hearing taking place next week about artificial intelligence, robotics and the fourth industrial revolution. “If we’ve got the march of the robots, we perhaps need the march of the robots to our select committee to give evidence,” Committee chair Robert Halfon told Tes. “The fourth industrial revolution is possibly the most important challenge facing our nation over the next 10, 20 to 30 years.” It is not clear whether Pepper will be pre-programmed to answer the questions or if it will rely on artificial intelligence to respond. The Independent has reached out for more details about the appearance.
2018-10-09 14:15:00+01:00 Read the full story.

 

Training machine learning models online for free (GPU, TPU enabled)!!!

Computation power needed to train machine learning and deep learning model on large datasets, has always been a huge hindrance for machine learning enthusiast. But with jupyter notebook which run on cloud anyone who is has the passion to learn can train and come up with great results. In this post I will providing information about the various service that gives us the computation power to us for training models.

  • Google Colab
  • Kaggel Kernel
  • Jupyter Notebook on GCP
  • Amazon SageMaker
  • Azure Notebooks

2018-10-12 14:54:23.385000+00:00 Read the full story.

 

Python Pandas vs. Scala: how to handle dataframes (part II)

A few days ago I published a post comparing the basic commands of Python and Scala: how to deal with lists and arrays, functions, loops, dictionaries and so on. As I continue practicing with Scala, it seemed appropriate to follow-up with a second part, comparing how to handle dataframes in the two programming languages, in order to get the data ready before the modeling process. In Python, we will do all this by using Pandas library, while in Scala we will use Spark.
2018-10-15 02:31:06.105000+00:00 Read the full story.

 

Celonis brings intelligent process automation software to cloud

Celonis has been helping companies analyze and improve their internal processes using machine learning. Today the company announced it was providing that same solution as a cloud service with a few nifty improvements you won’t find on prem. The new approach, called Celonis Intelligent Business Cloud, allows customers to analyze a workflow, find inefficiencies and offer improvements very quickly. Companies typically follow a workflow that has developed over time and very rarely think about why it developed the way it did, or how to fix it. If they do, it usually involves bringing in consultants to help. Celonis puts software and machine learning to bear on the problem. Co-founder and CEO Alexander Rinke says that his company deals with massive volumes of data and moving all of that to the cloud makes sense. “With Intelligent Business Cloud, we will unlock that [on prem data], bring it to the cloud in a very efficient infrastructure and provide much more value on top of it,” he told TechCrunch.
2018-10-15 00:00:00 Read the full story.

 

10 Best Mobile Apps for Data Scientist / Data Analysts

Data science and machine learning are evolving with their abilities to transform the world around you. You don’t need carry your laptop or PC 24/7 with you to excel in your workplace. It is time to take a break from daily routines and adopt to quicker learning strategies. The solution is simple, just switch to Mobile Apps. Did you know you can run Python in your Smartphone? Yeah! Mobile applications have had their own share and added immense excitement to our methods for learning. The subjects which were hard to understand, are currently taught utilizing pictures and stories in your mobile or tablet. You can retrieve them anyplace, even if you are travelling or not.

In this blog, we have shared some valuable apps which can enhance your vital data science and analytics skills. These applications can enhance your listening abilities, logical skills, basic leadership qualities etc., which tends to be more powerful than our expectation. We have assembled these mobile applications in various categories, we all are aware of our weaknesses, so this blog would assist you with targeting the sweet spots. These android applications are available for free at google play store.

  • Elevate (Downloads – 1 million)
  • Lumosity (Downloads – 10 million)
  • Neuronation (Downloads – 5 million)
  • Math Workout (Downloads – 5 million)
  • QPython (Downloads – 500k)
  • Learn Python (Downloads – 10k)
  • R Programming (Downloads – 10k)
  • Basic Statistics (Downloads – 50k)
  • Probability Distributions (Downloads – 50k)
  • Udacity (Downloads – 1 million)

2018-10-10 00:00:00 Read the full story.

 

Q&A: Why Reserved Instance optimisation is the future of cloud cost management

Why is cloud cost management becoming such a headache for enterprise customers? Cloud operations teams are stuck between a finance department seeking to drive down cloud costs and application owners demanding more cloud resources. On top of that, matching workload demands with the right public cloud products is complex: demand patterns are hard to identify and there are simply too many cloud service combinations to choose from.
2018-10-15 00:00:00 Read the full story.

 

The top pitches from Amazon’s Alexa Accelerator as startups push boundaries of voice tech

Nine startups from across the world spent the past three months as part of the second cohort of the Alexa Accelerator, a Seattle-based program co-led by Techstars and Amazon.
Entrepreneurs built out B2C and B2B technologies that incorporate Alexa, Amazon’s artificial intelligence and machine learning-powered voice platform.
2018-10-10 17:52:59-07:00 Read the full story.

 

Voice-first, AI, AR: Three technologies race to again change how we shop

Ecommerce is known for its waves of innovation. From the rise of pure-plays to the advent of multi-channel and subscription commerce, we’ve seen the category disrupted more than once. Now, a handful of innovations are vying to again change how we buy as these technologies move from the emergent to the mainstream. Augmented Reality (AR) got off to a slower-than-expected start but appears to be gaining steam. Artificial Intelligence (AI) is rapidly shedding its dystopian baggage to find a home in countless applications. And voice-first design for smart devices? It’s a thing. Each of the three is affecting commerce in its own unique way: one more behind the scenes, the others far more in (or literally on) your face.
2018-10-13 14:00:40-07:00 Read the full story.

 

Siamese Networks and Stuart Weitzman Boots – Towards Data Science

In a previous post I wrote about how you could use a Mask R-CNN model to detect and segment out articles of clothing to be used by some second stage model. For his post I built an example of that second stage model using a Pytorch siamese neural network. The idea would be that by combining these two models you could take a raw image, segment out just the articles of clothing, and then match those articles of clothing against a database of clothing items to find similar items.

So the first part of this post will focus on building the siamese network and near the end I will show an example using the output from my segmentation model and how the siamese network can generalize beyond shoes to other items it has never seen before.
2018-10-15 13:32:32.621000+00:00 Read the full story.

 

Apple bought a music analytics company that’s likely to beef up Apple Music playlists

Apple has bought music analytics company Asaii, according to a LinkedIn profile update from the start-ups co-founder, in a move that’s likely to make Apple Music playlists better. Asaii ranks tracks and artists by popularity across music streaming platforms. The company’s machine learning algorithms predict which artists will top the charts next — “to find the next Justin Bieber, before anyone else,” the company says. It’s another means to bolster Apple Music as the company takes on streaming giant Spotify. Apple bought music recognition app Shazam earlier this year.
2018-10-15 00:00:00 Read the full story.

 


Behind a Paywall/Registration Wall…

There’s a mismatch at big US investment firms on the importance of AI, and it could highlight a level of ‘complacency’

Everyone’s talking about artificial intelligence — but big US investment funds aren’t yet keen to try it out, according to a new study.

About 71% of US-based firms are not currently testing or considering how AI and advanced analytics can be applied to their investments, said a Fidelity survey of over 900 institutional investors published on Thursday. However, a similar percentage of US investors agreed that AI and technological advances will …
2018-10-11 00:00:00 Read the full story.
1.6941176470588233

 


This news clip post is produced algorithmically based upon CloudQuant’s list of sites and focus items we find interesting. 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.