Cloud Hosting

Its no secret that creating and deploying machine learning models takes too long. In Algorithmias 2021 Enterprise Trends in Machine Learning, 25% of respondents said creating a model took one week to one month, while 24% put that time at one month to one quarter. And 37% said it took one quarter to one year to deploy a model.

At Uber, an intent classification system that involved 1,500 lines of TensorFlow took five months to create and seven months to deploy.

A second machine learning project, fraud detection, with 900 lines of PyTorch, took five months to create and four months to deploy.

A product recommendation tool with 1,200 lines of PyTorch took six months to create and seven months to deploy.

San Francisco-based startup Predibase is out to change that by providing a low-code declarative ML platform that both data scientists and non-experts can use, easing the pressure on organizations to hire more scarce and expensive data scientists. Users can just state what they want to do starting with just six lines of Python code and let the system figure out how to do it and the infrastructure required.

Its built atop two machine learning technologies created by the Predibase founders at Uber: Ludwig and Horovod. Ludwig is an open source,declarative machine learning frameworkthat provides the simplicity of an autoML solution with the flexibility of writing your own PyTorch code. Horovod, an open source component of Ubers Michelangelo deep learning toolkit makes it easier to start and speed up distributed deep learning projects with TensorFlow.

The experience is that data science organizations have to basically reinvent the wheel and create a bespoke solution for every single one of these products, and theres not much in common among them. Because of that, the whole organization becomes a bottleneck for machine learning adoption, said Predibase CEO, Piero Molino. The result is that it just takes too long for machine learning models to bring value to an organization.

In contrast, he compares a declarative configuration system to what Kubernetes has done for infrastructure.

Our vision is to make machine learning as easy as writing a SQL query, Molino said.

The basic idea is to let users specify entire model pipelines as configurations the parts they care about and automate the rest.

Traditional machine learning projects involve a complicated ML life cycle that spans feature and data engineering; model development and training; and model production and governance. Cross-functional data science teams struggle to manage these phases in a coherent and sustainable way, said Kevin Petrie, vice president of research at Eckerson Group.

Predibase represents a level of innovation to simplify the ML life cycle. Predibase proposes to let data science teams specify the desired inputs and outputs for their ML model. That is, they create configuration files that Predibase then figures out how to implement. Data science teams still can customize as many parameters, etc. as they like by making modular changes to meet new or changing customer requirements.

In short, Predibase proposes to minimize the complexity of the ML life cycle, which is the biggest barrier to success with data science projects.

Its easy to get started. That Uber intent classification system could be created, for example with six lines of code. You get something that is readable and reproducible and shareable, he said.

But one of the advantages is that you retain all the flexibility and control that an expert needs. So you can specify through the configuration all the details about the models choosing among different model architectures, training parameters, about the preprocessing of the data. Its all accessible through a parameter in the configuration, which makes it easy to iterate and improve models. Make changes with just a new configuration.

Its also extensible. So if youre an expert developer, you can add your own keys to the configuration. You can extend this by adding your own piece of PyTorch, for instance, and then it can be referenced from the configuration.

The company has deep expertise in machine learning.

Molino, the creator of Ludwig, previously was staff research scientist at Stanford University and co-founder and senior research scientist at Uber AI.

Fellow Predibase co-founders are:

Predibase enables users to easily connect to structured and unstructured data stored anywhere on the cloud data stack; write model pipeline configurations and run on a scalable distributed infrastructure to train models as easily as on a single machine; deploy model pipelines with the click of a button and query them immediately.

Predibase is building the first declarative ML platform that enables enterprises to develop and operationalize models, from data to deployment, without having to choose between simplicity and the power of fine-grained controls. The rapid success of both the open source foundations and the beta of its commercial platform in the Fortune 500 has been incredibly exciting, Greylock Partner Saam Motamedi said at the recent announcement of a $16.25 million Series A round.

Still in private beta with Fortune 500 customers, Predibase is looking toward a general release in the second half of this year.

Customers have been using datasets of about 1 billion to 2 billion rows about 100 to 200 columns and several hundred gigabytes. Internal benchmarking has run up to 2 terabytes. Ludwig and Horovod, however, have been tested on much larger data set sizes even than that, according to Rishi.

The company maintains it takes a different approach than other automated machine learning products.

Thinking of something like DataRobot or Google Cloud AutoML, for example, [they] provide these interfaces where you kind of bring in data, click a button and you get models out, explained Molino. We found that thats actually pretty unsatisfying for a lot of users and customers because they tend to be black boxes that dont have any configurability or control. So the minute that the platform doesnt give you a good out-of-the-box model, youre kind of stuck, and you end up graduating out.

Users can access the capabilities in Predibase purely through Python, through the UI or through PQL (Predictive Query language), an extension of SQL.

The PQL extension includes predicates that allow you to bring machine learning and data together, Rishi explained. Its flexibility puts machine learning in the language of, of data users, so they can use filter, group by aggregate, join or any other commands that theyre familiar with in SQL. Its extensible. Simply add new features as an additional predicate. Predibase makes it just as easy to use text and image and other types of fields as standard tabular fields.

This is really simple. It brings machine learning into the hands of a broader set of users that are familiar with SQL, but at the same time, behind the scenes, the power and flexibility of the Ludwig configuration system provide state-of-the-art performance on both structured and unstructured data, and the combination of the two, Molino said.

And finally, we also abstract away the infrastructure based on Horovod, they can train and deploy models at scale. And its basically a big-tech-level infrastructure without the need to have a big-tech-level engineering team to build it, right. Its already built for you.

Models can be queried as REST APIs, through the Python SDK and through the PQL language. Though the entire process is encapsulated in the platform, the models also can be exported, should the user need to run them elsewhere.

A model repositories page summarizes the models just as configuration, making comparing model versions easy.

The company is spending the first half of this year making that product enterprise-ready with robustness, enterprise-grade security and enabling multicloud deployments, Molino said. After a GA launch, it wants to pursue integrations with the wider ML ecosystem, with tools like dbt, for instance, and eventually make Predibase self-service.

Feature image via Pixabay

Read this article:

Predibase Takes Declarative Approach to AutoML The New Stack - thenewstack.io

As Aotearoa New Zealand heads into the colder winter months, the pressures on our health system and staff are growing significantly.

On top of the ongoing impact of COVID-19, flu cases have begun to spike.

Conditions are also primed for potential outbreaks of other illnesses including measles, whooping cough and respiratory syncytial virus (RSV).

If we are to weather the coming storm, there will need to be a recommitment to public health measures that slow the spread of respiratory infections, as well as a renewed drive for widespread vaccination.

The first wave of Omicron swept through Aotearoa New Zealand in late February and March.

Unfortunately, as seen in many other countries, the fall in case numbers has been much slower than the rise, with infections reaching a plateau in all age groups.

Case numbers have been driven by a high number of infections in young people between the ages of 10 and 29 years old. But the elderly have borne the brunt of hospitalisations, largely due to the higher risk of severe outcomes for older adults.

Age stratification aside, persistent inequities have also left Mori and Pasifika at the sharp end of the outbreak both in terms of cases and severe outcomes.

Hospitalisation rates and reinfections are rising in many age groups, mirroring trends seen elsewhere.

New Zealand can expect another resurgence of COVID-19 this winter.

While 95% of New Zealand has received the second dose of the vaccine, one of the highest rates in the world, fewer have received a booster. We also have lower than optimal levels of childhood vaccination.

Long COVID will add a layer of complication for our medical services.

A recent report by the US Center for Disease Control (CDC) suggests one in five COVID-19 survivors aged 18 to 64 years old, and one in four survivors aged 65 years and above, experienced at least one condition that might be attributable to previous COVID-19 infection.

Despite being labelled as one of this generations disability challenges, there is currently no test for long COVID.

Worryingly, COVID-19 deaths in Australia have started to trend upwards. Evidence from Australia has shown that the overwhelming majority of people are dying from, not with, COVID-19.

Health professionals are not just worried about COVID-19. The flu and other viruses are also expected to hit hard this year.

Thanks to closed borders, managed isolation and quarantine, and lockdowns, the last time New Zealand experienced a flu season was in 2019.

We are now more vulnerable to the virus. There has already been a reported surge in Dunedin.

In response, the government has made two million vaccines available and has the widened eligibility for people to get vaccinated for free.

Unfortunately, there is growing concern that part of the population may not get vaccinated due to immunisation fatigue, or may be unable to due to structural inequities in access to vaccines.

As with COVID-19, looking across the Tasman can help us understand what is likely to happen in New Zealand.

Much like New Zealand, flu rates in Australia have, until now, been very low due to closed borders.

The latest Australian national surveillance for influenza shows a steep rise in rates of the flu, as well as rising hospital and ICU admissions.

Before the COVID-19 pandemic even started, our research highlighted declines in childhood immunisation for vaccine-preventable diseases.

Public health officials are now noticing further significant declines in routine childhood immunisations.

In April, the World Health Organization reported a 79% increase in measles cases in the first two months of 2022.

Meaningfully addressing long-standing inequities in childhood vaccination programmes takes on new urgency in the face of these vaccination gaps.

Lessons can also be learned from the COVID-19 vaccination programme regarding the success of handing leadership to Mori and Pasifika community providers to improve vaccination rates.

We have long been warned that an underfunded health system might struggle with a seasonal surge in demand.

Pressure points have appeared across the country. On May 23, Dunedin Hospitals COVID-19 ward was at capacity. Two days later, Nelson Hospital also hit capacity limits, leading to temporary ambulance ramping at the emergency department.

Canterbury District Health Board, Hawkes Bay District Health Board, and MidCentral District Health Board have recently urged people to consider alternative care for minor conditions to help alleviate the pressure.

Community health providers are also struggling to meet demands.

During the winter, we spend more time in indoor spaces with inadequate ventilation. We are also becoming more complacent with our mask wearing as policies relax.

In the future, vaccines will need to improve.

But for now, its important to remember that three doses of the COVID-19 vaccine remain effective against hospitalisation even for newer variants, as well as lowering the risk of infection.

But there are things we can all do to avoid the worst this winter has to offer, including to:

Finally, workplaces should continue to support people to stay home and isolate if required.

ABOUT THE AUTHORS

Matthew Hobbs, Senior Lecturer in Public Health and Co-Director of the GeoHealth Laboratory, University of Canterbury. As a researcher he is interested in modifying environments to improve behaviour/health. His research uses geospatial and population health methods to inform local and national public health policy to reduce health inequities. Most recently he was awarded by the European Association for the Study of Obesity (2019) the New Investigator in Public Health Award which recognizes rising stars in the field of obesity through their wider contributions to public health and it is highly regarded throughout the scientific community.

Alex Kazemi, Intensive Care specialist, University of Auckland. Dr. Alex Kazemi has a background in clinical medicine, but is currently pursuing a Masters degree in Public Health at the University of Auckland. He completed his MBBS at Kings College, London, graduating as a doctor in 1996 from Kings College Medical School. Subsequently he completed specialist training in Aotearoa with fellowships in Emergency Medicine and Intensive Care. He worked as an Intensive Care Specialist at Middlemore Hospital, Auckland for ten years. He has interests in use of data analytics for public health and health surveillance as well as in the communication of public health data.

Lukas Marek, Researcher and lecturer in Spatial Data Science, University of Canterbury. I received my PhD in 2015 from the Department of Geoinformatics at Palacky University in the Czech Republic. In my thesis, I conducted a nationwide research study on the spatial and spatiotemporal analysis of gastroenteritis. My main research interest is mostly focused on spatial analyses and (geo)visualisation of health- and wellbeing-related data using spatial data science.

This article is courtesy of The Conversation.

See more here:

Resurgent COVID-19, Flu and Other Viruses are Pushing New Zealand's Health System to the Limit - Global Biodefense

The annual median pay for a career in data analytics can range from $82,000 to $114,000. Additionally, market-related analytics is one of the fastest growing fields with Forbes estimating a growth rate of 67% over the past year, and 136% over the past three years and the U.S. Bureau of Labor Statistics projecting employment growth of up to 27% from 2016 to 2026.

The University of Scrantons Bachelor of Science in business analytics is designed to help students learn the variety of skill sets needed to be successful in this interdisciplinary field, where they will be asked to conduct, analyze and evaluate data in order to make better business decisions. Potential job titles one can pursue with this degree include management consultant, data analyst, operations research analyst, and more.

As part of the Kania School of Management (KSOM), Scrantons business programs are accredited by the rigorous standards of the Association to Advance Collegiate Schools of Business (AACSB), which is an elite accolade that fewer than five percent of business schools worldwide hold.

Scrantons business analytics program includes courses from data mining, database management systems, and simulation as well as the communication, managerial and decision-making skills needed to Students will be introduced to tools like structured query language (SQL) and Visual Basic for Applications (VBA) as related to the use of analytics in decision making.

Scranton also offers a masters degree in business analytics in an online or on-campus format, as well as a business analytics certificate.

Continue reading here:

Predicting the Future with Business Analytics | Royal News: June 1 2022 - Scranton

Andreas Braun is the Director Artificial Intelligence & Data Science at PwC Luxembourg. He is passionate about technologies that have the human factor at the forefront. Biometrics, in particular in its usage of Machine Learning and AI methods is his area of expertise. We caught up with him to find out how his role has developed, and why AI is at the forefront of the most innovative fintech solutions.

I am the AI lead at PwC Luxembourg, which gives me the great opportunity to have an impact on all the different sectors that are touched by AI which pretty much means all of them.

While Luxembourg is first of all known as a prime destination for the fund industry, it also has a very healthy start-up scene, with a strong focus on fintech. We are involved as mentors and partners for many of the initiatives in this area. AI plays a key role in many innovative products and services provided by fintech companies.

The last decade in AI has been truly exceptional. From humble attempts to improve scanning documents, Deep Learning and further developments have been truly exciting. It enabled computers to overtake machines in many tasks, from object recognition, face detection, to many predictive tasks that are highly relevant in the financial sector.

There are however limitations for the foreseeable future. Just like our cars are not fully self-driving yet, we still need the human factor in many investment and financial decisions and for the foreseeable future this will not change. There is a lot of opportunity in combining human ingenuity and AIs capacity in learning rapidly from large amounts of data.

Pretty much all fintechs use data, and most of them AI. So let me just pick two examples. Know Your Customer (KYC) and Anti Money Laundering (AML) are areas where AI-powered robots are able to rapidly screen massive amounts of information and find links between globally distributed information sources. Imagine a fintech that would have predicted (and maybe prevented) the Wirecard scandal if it happened just a few years later.

RegTech is exploding using AI technologies to deal with the increasing cost and complexity of compliance. AI systems assess the most recent regulations and legal communications from governments around the world and compare it with your procedures. The cost savings can be massive. And of course, fintechs could help regulators create more effective regulation, too.

Maybe lets just focus on one key regulation, the European AI Act that has been in the making for almost two years. It foresees a risk-based approach towards the use of AI, when dealing with customer data, or when making decisions that may impact your customers, which covers a large chunk of AI applications. Credit risk for persons and claims processing are two areas in which many fintechs are active that are specifically mentioned as high risk. However, the regulation is still under discussion, so changes may still apply in the future.

The smart incumbent banks are looking at inspiration or see it as a potential investment opportunity. In terms of personal banking, many have rolled out features that leverage AI to analyse personal spending habits, also in order to suggest suitable products. Others acquire teams in order to foster their own innovation capacity.

The sector is however still hampered by a long legacy of data processing and storage, meaning that they are partially working with very old infrastructures, dissociated data sources and manual processes that prevent the effective use of data and in turn, AI.

Looking into the crystal ball is always challenging, but one topic will be hyper personalisation of services, which becomes more feasible with advanced AI techniques. The capacity to provide your client with the perfect service is bound to be a significant competitive advantage. This has driven the success of many shopping and streaming services and the AI investments and advances are massive. A fintech that cracks this case will have a chance to become the next unicorn.

One can see this as a continuation of existing trends. Mobile first, multi-channel services have been shaping the financial industry already over the past years. These technologies will allow our mobile devices to become more powerful and our access to online financial services becoming more immersive. Personally, I am not sure if I want to check how my stocks are doing through VR goggles, but it will enable more services being available 24/7. Honestly, who really wants to spend 2 hours in traffic and the waiting room, to get a single signature from your agent?

Curiosity has been driving me in both my research and professional career. The fintech sector has an exceptional capacity to pique my curiosity. I never cease to be amazed by the new means and ways in which fintech companies are using the latest innovations from research, find new applications for old stuff and come up with amazing solutions to many of our main pressing global challenges.

See the original post here:

Why AI and ML is Reshaping the Fintech Industry - FinTech Magazine