Mediaboss Marketing

Yi-ming Li,1,* Li-cheng Jiang,2,* Jing-jing He,1 Kai-yu Jia,1 Yong Peng,1 Mao Chen1

1Department of Cardiology, West China Hospital, Sichuan University, Chengdu, Peoples Republic of China; 2Department of Cardiology, The First Affiliated Hospital, Chengdu Medical College, Chengdu, Peoples Republic of China

*These authors contributed equally to this work

Correspondence: Yong Peng; Mao ChenDepartment of Cardiology, West China Hospital, Sichuan University, 37 Guoxue Street, Chengdu 610041, Peoples Republic of ChinaEmail pengyongcd@126.com; hmaochen@vip.sina.com

Abstract: A formal risk assessment for identifying high-risk patients is essential in clinical practice and promoted in guidelines for the management of anterior acute myocardial infarction. In this study, we sought to evaluate the performance of different machine learning models in predicting the 1-year mortality rate of anterior ST-segment elevation myocardial infarction (STEMI) patients and to compare the utility of these models to the conventional Global Registry of Acute Coronary Events (GRACE) risk scores. We enrolled all of the patients aged >18 years with discharge diagnoses of anterior STEMI in the Western China Hospital, Sichuan University, from January 2011 to January 2017. A total of 1244 patients were included in this study. The mean patient age was 63.812.9 years, and the proportion of males was 78.4%. The majority (75.18%) received revascularization therapy. In the prediction of the 1-year mortality rate, the areas under the curve (AUCs) of the receiver operating characteristic curves (ROCs) of the six models ranged from 0.709 to 0.942. Among all models, XGBoost achieved the highest accuracy (92%), specificity (99%) and f1 score (0.72) for predictions with the full variable model. After feature selection, XGBoost still obtained the highest accuracy (93%), specificity (99%) and f1 score (0.73). In conclusion, machine learning algorithms can accurately predict the rate of death after a 1-year follow-up of anterior STEMI, especially the XGBoost model.

Keywords: machine learning, prediction model, acute anterior myocardial infarction

This work is published and licensed by Dove Medical Press Limited. The full terms of this license are available at https://www.dovepress.com/terms.php and incorporate the Creative Commons Attribution - Non Commercial (unported, v3.0) License.By accessing the work you hereby accept the Terms. Non-commercial uses of the work are permitted without any further permission from Dove Medical Press Limited, provided the work is properly attributed. For permission for commercial use of this work, please see paragraphs 4.2 and 5 of our Terms.

Read the original:
Machine Learning to Predict the 1-Year Mortality Rate After Acute Ante | TCRM - Dove Medical Press

MIT boffins have devised a software-based tool for predicting how processors will perform when executing code for specific applications.

In three papers released over the past seven months, ten computer scientists describe Ithemal (Instruction THroughput Estimator using MAchine Learning), a tool for predicting the number processor clock cycles necessary to execute an instruction sequence when looped in steady state, and include a supporting benchmark and algorithm.

Throughput stats matter to compiler designers and performance engineers, but it isn't practical to make such measurements on-demand, according to MIT computer scientists Saman Amarasinghe, Eric Atkinson, Ajay Brahmakshatriya, Michael Carbin, Yishen Chen, Charith Mendis, Yewen Pu, Alex Renda, Ondrej Sykora, and Cambridge Yang.

So most systems rely on analytical models for their predictions. LLVM offers a command-line tool called llvm-mca that can presents a model for throughput estimation, and Intel offers a closed-source machine code analyzer called IACA (Intel Architecture Code Analyzer), which takes advantage of the company's internal knowledge about its processors.

Michael Carbin, a co-author of the research and an assistant professor and AI researcher at MIT, told the MIT News Service on Monday that performance model design is something of a black art, made more difficult by Intel's omission of certain proprietary details from its processor documentation.

The Ithemal paper [PDF], presented in June at the International Conference on Machine Learning, explains that these hand-crafted models tend to be an order of magnitude faster than measuring basic block throughput sequences of instructions without branches or jumps. But building these models is a tedious, manual process that's prone to errors, particularly when processor details aren't entirely disclosed.

Using a neural network, Ithemal can learn to predict throughout using a set of labelled data. It relies on what the researchers describe as "a hierarchical multiscale recurrent neural network" to create its prediction model.

"We show that Ithemals learned model is significantly more accurate than the analytical models, dropping the mean absolute percent error by more than 50 per cent across all benchmarks, while still delivering fast estimation speeds," the paper explains.

A second paper presented in November at the IEEE International Symposium on Workload Characterization, "BHive: A Benchmark Suite and Measurement Framework for Validating x86-64 Basic Block Performance Models," describes the BHive benchmark for evaluating Ithemal and competing models, IACAm llvm-mca, and OSACA (Open Source Architecture Code Analyzer). It found Ithemal outperformed other models except on vectorized basic blocks.

And in December at the NeurIPS conference, the boffins presented a third paper titled Compiler Auto-Vectorization with Imitation Learning that describes a way to automatically generate compiler optimizations in a way that outperforms LLVMs SLP vectorizer.

The academics argue that their work shows the value of machine learning in the context of performance analysis.

"Ithemal demonstrates that future compilation and performance engineering tools can be augmented with datadriven approaches to improve their performance and portability, while minimizing developer effort," the paper concludes.

Read the original here:
Finally, a good use for AI: Machine-learning tool guesstimates how well your code will run on a CPU core - The Register

Every hotel should ask the same question. How will our property use machine learning? Its not just a matter of gaining a competitive advantage; its imperative in order to stay in business.By Jason G. Bryant, Founder and CEO, Nor1 - 1.9.2020

Artificial intelligence (AI) implementation has grown 270% over the past four years and 37% in the past year alone, according to Gartners 2019 CIO Survey of more than 3,000 executives. About the ubiquity of AI and machine learning (ML) Gartner VP Chris Howard notes, If you are a CIO and your organization doesnt use AI, chances are high that your competitors do and this should be a concern, (VentureBeat). Hotels may not have CIOs, but any business not seriously considering the implications of ML throughout the organization will find itself in multiple binds, from the inability to offer next-level guest service to operational inefficiencies.

Amazon is the poster child for a sophisticated company that is committed to machine learning both in offers (personalized commerce) as well as behind the scenes in their facilities. Amazon Founder & CEO Jeff Bezos attributes much of Amazons ongoing financial success and competitive dominance to machine learning. Further, he has suggested that the entire future of the company rests on how well it uses AI. However, as Forbes contributor Kathleen Walsh notes, There is no single AI group at Amazon. Rather, every team is responsible for finding ways to utilize AI and ML in their work. It is common knowledge that all senior executives at Amazon plan, write, and adhere to a six-page business plan. A piece of every business plan for every business function is devoted to answering the question: How will you utilize machine learning this year?

Every hotel should ask the same question. How will our property use machine learning? Its not just a matter of gaining a competitive advantage; its imperative in order to stay in business. In the 2017 Deloitte State of Cognitive Survey, which canvassed 1,500 mostly C-level executives, not a single survey respondent believed that cognitive technologies would not drive substantive change. Put more simply: every executive in every industry knows that AI is fundamentally changing the way we do business, both in services/products as well as operations. Further, 94% reported that artificial intelligence would substantially transform their companies within five years, most believing the transformation would occur by 2020.

Playing catch-up with this technology can be competitively dangerous as there is significant time between outward-facing results (when you realize your competition is outperforming you) and how long it will take you to achieve similar results and employ a productive, successful strategy. Certainly, revenue management and pricing will be optimized by ML, but operations, guest service, maintenance, loyalty, development, energy usage, and almost every single aspect of the hospitality enterprise will be impacted as well. Any facility where the speed and precision of tactical decision making can be improved will be positively impacted.

Hotels are quick to think that when ML means robotic housekeepers and facial recognition kiosks. While these are possibilities, ML can do so much more. Here are just a few of the ways hotels are using AI to save money, improve service, and become more efficient.

Hiltons Energy Program

The LightStay program at Hilton predicts energy, water, and waste usage and costs. The company can track actual consumption against predictive models, which allows them to manage year-over-year performance as well as performance against competitors. Further, some hotel brands can link in-room energy to the PMS so that when a room is empty, the air conditioner automatically turns off. The future of sustainability in the hospitality industry relies on ML to shave every bit off of energy usage and budget. For brands with hundreds and thousands of properties, every dollar saved on energy can affect the bottom line in a big way.

IHG & Human Resources

IHG employs 400,000 people across 5,723 hotels. Holding fast to the idea that the ideal guest experience begins with staff, IHG implemented AI strategies tofind the right team member who would best align and fit with each of the distinct brand personalities, notes Hazel Hogben, Head of HR, Hotel Operations, IHG Europe. To create brand personas and algorithms, IHG assessed its top customer-facing senior managers across brands using cognitive, emotional, and personality assessments. They then correlated this with KPI and customer data. Finally, this was cross-referenced with values at the different brands. The algorithms are used to create assessments to test candidates for hire against the personas using gamification-based tools, according to The People Space. Hogben notes that in addition to improving the candidate experience (they like the gamification of the experience), it has also helped in eliminating personal or preconceived bias among recruiters. Regarding ML uses for hiring, Harvard Business Review says in addition to combatting human bias by automatically flagging biased language in job descriptions, ML also identifies highly qualified candidates who might have been overlooked because they didnt fit traditional expectations.

Accor Hotels Upgrades

A 2018 study showed that 70% of hotels say they never or only sometimes promote upgrades or upsells at check-in (PhocusWire). In an effort to maximize the value of premium inventory and increase guest satisfaction, Accor Hotels partnered with Nor1 to implement eStandby Upgrade. With the ML-powered technology, Accor Hotels offers guests personalized upgrades based on previous guest behavior at a price that the guest has shown a demonstrated willingness to pay at booking and during the pre-arrival period, up to 24 hours before check-in. This allows the brand to monetize and leverage room features that cant otherwise be captured by standard room category definitions and to optimize the allocation of inventory available on the day of arrival. ML technology can create offers at any point during the guest pathway, including the front desk. Rather than replacing agents as some hotels fear, it helps them make better, quicker decisions about what to offer guests.

Understanding Travel Reviews

The luxury Dorchester Collection wanted to understand what makes their high-end guests tick. Instead of using the traditional secret shopper methods, which dont tell hotels everything they need to know about their experience, Dorchester Collection opted to analyze traveler feedback from across major review sites using ML. Much to their surprise, they discovered Dorchesters guests care a great deal more about breakfast than they thought. They also learned that guests want to customize breakfast, so they removed the breakfast menu and allowed guests to order whatever they like. As it turns out, guests love this.

In his May 2019 Google I/O Address, Google CEO Sundar Pichai said, Thanks to advances in AI, Google is moving beyond its core mission of organizing the worlds information. We are moving from a company that helps you find answers to a company that helps you get things done (ZDNet). Pichai has long held that we no longer live in a mobile-first world; we now inhabit an AI-first world. Businesses must necessarily pivot with this shift, evolving processes and products, sometimes evolving the business model, as in Googles case.

Hotels that embrace ML across operations will find that the technologies improve processes in substantive ways. ML improves the guest experience and increases revenue with precision decisioning and analysis across finance, human resources, marketing, pricing and merchandising, and guest services. Though the Hiltons, Marriotts, and IHGs of the hotel world are at the forefront of adoption, ML technologies are accessibleboth in price and implementationfor the full range of properties. The time has come to ask every hotel department: How will you use AI this year?

For more about Machine Learning and the impact on the hotel industry, download NOR1s ebook The Hospitality Executives Guide to Machine Learning: Will You Be a Leader, Follower, or Dinosaur?

Jason G. Bryant, Nor1 Founder and CEO, oversees day-to-day operations, provides visionary leadership and strategic direction for the upsell technology company. With Jason at the helm, Nor1 has matured into the technology leader in upsell solutions. Headquartered in Silicon Valley, Nor1 provides innovative revenue enhancement solutions to the hospitality industry that focus on the intersection of machine learning, guest engagement and operational efficiency. A seasoned entrepreneur, Jason has over 25 years experience building and leading international software development and operations organizations.

Related

Read the rest here:
How Will Your Hotel Property Use Machine Learning in 2020 and Beyond? | - Hotel Technology News

We tend to think that the lowest point of entry for machine learning (ML) is on a Raspberry Pi, which it definitely is not. [EloquentArduino] has been pushing the limits to the low end of the scale, and managed to get a basic classification model running on the ATtiny85.

Using his experience of running ML models on an old Arduino Nano, he had created a generator that can export C code from a scikit-learn. He tried using this generator to compile a support-vector colour classifier for the ATtiny85, but ran into a problem with the Arduino ATtiny85 compiler not supporting a variadic function used by the generator. Fortunately he had already experimented with an alternative approach that uses a non-variadic function, so he was able to dust that off and get it working. The classifier accepts inputs from an RGB sensor to identify a set of objects by colour. The model ended up easily fitting into the capabilities of the diminutive ATtiny85, using only 41% of the available flash and 4% of the available ram.

Its important to note what [EloquentArduino] isnt doing here: running an artificial neural network. Theyre just too inefficient in terms of memory and computation time to fit on an ATtiny. But neural nets arent the only game in town, and if your task is classifying something based on a few inputs, like reading a gesture from accelerometer data, or naming a color from a color sensor, the approach here will serve you well. We wonder if this wouldnt be a good solution to the pesky problem of identifying bats by their calls.

We really like how approachable machine learning has become and if youre keen to give ML a go, have a look at the rest of the EloquentArduino blog, its a small goldmine.

Were getting more and more machine learning related hacks, like basic ML on an Arduino Uno, and Lego sortings using ML on a Raspberry Pi.

The rest is here:
Tiny Machine Learning On The Attiny85 - Hackaday

Suppose you are driving a hybrid car with a personalised Alexa prototype and happen to witness a road accident. Will your Alexa automatically stop the car to help the victim or call an ambulance? Probably,it would act according tothe algorithmprogrammed into itthat demands the users command.

But as a fellow traveller with Alexa, what would you do? If you areanempathetic human being, you would try to administer first aid and take the victim to a nearby hospital in your car. This empathy is what is missing in the machines, largely in the technocratic conquered education which parents are banking upon these days.

Tech-buddies

With the advancement of bots or robots teaching in our classrooms, theteachersof millennials are worried. Recently, a WhatsApp video of AI-teacher engaging class in one of the schools of Bengaluru went viral. Maybe in a decade or two, academic robots in our classrooms would teach mathematics. Or perhaps they will teach children the algorithmsthatbrings them to life and togetherthey can create another generation of tech-buddies.

I was informed by a friend that coding is taught atprimary level now which was indeed a surprise for me. Then what about other skills? Maybe life skills like swimming, cooking could also be taught by a combination of YouTube and personal robots. However, we have the edge over the machines in at least one area and thats basic human values. This is where human intervention cant be eliminated at all.

The values are not taught; rather they are ingrained at every phase of life by various people who we meet including parents, teachers, peers, and anyone around us alongside practising them. Say for example, how does one teach kids to care for the elderly at home?

Unless one feels the same emotional turmoilas the elderly before them as they are raised and apply the compassionate values, they wouldnt be motivated to take care of them.

The missing link in academia

The discussions on trans-disciplinary or interdisciplinary courses often put forward multiple subjects as well as unconventional subjects to study together. Like engineering and terracotta designs or literature and agriculture. However, the objection comes within academia citing a lack of career prospects.

We tend to forget the fact that the best mathematicians were also musicians and the best medicinal practitioners were botanists or farmers too. Interest in one subject might trigger gaining expertise in others and connect the discreet dots to create a completely new concept.

Life skills like agriculture, pottery, animal care, gardening, andhousing are essentialskills that have many benefits.Every rural person is equipped with these skills through surrounding experiences. Rather than in a classroom session, these learning takes place by seeing, interacting as well as making mistakes.

A friend who homeschooled both her kids had similar concerns. She was firmly against the formalised education which teaches a limited amount of information mostly based on memorisation taking out the natural interest of the child. Several such institutes are functioning to serve the same goals of lifelong learning. Such schools aiming at understanding human-nature, emotional wellbeing, artistic and critical thinking are fundamentally guided on the idea of learning in a fear-free environment.

When scrolling on the admissions page in these schools, I was surprised that the admissions for the 2021 academic year were already completed.This reflects the eagerness of many parents looking for such alternative education systems.

These analogies bring back the basic question of why education? If it is merely for technology-driven jobs, probably by the time your kids grow there wouldnt be many jobs as themachines would have snatched them.

Also, the country is moving towards a technology-driven economy and may not need many skilled labourers. Surely, a few post-millennials would survive in any condition if they are extremely smart and adoptive butthey may need to stop and reboot if theireducation has not prepared them for uncertainties to come.

(The writer is with Christ, Bengaluru)

Link:
Limits of machine learning - Deccan Herald