Cloud Hosting

Mastercards Ngoc Minh Tran discusses his role as lead data scientist and the growing importance of model explainability in AI and analytics.

Ngoc Minh Tran is a lead data scientist at Mastercard. Minh Tran began his career with a PhD in applied statistics/machine learning and has over 15 years of experience working in data science, with a focus on classical machine learning, deep learning and reinforcement learning.

With an impressive portfolio of achievements including being the lead inventor on 10 US patents in machine learning domains and the co-inventor of nine more we asked Minh Tran about what a typical day is like for a lead data scientist.

The most important thing for success in AI work is to practise as much as possible on multiple types of data

I always start a day with planning to work through the list of tasks under my responsibility, together with a target result or outcome. I also like to think about what tech book I am going to read if I manage to have some free time.

As a data scientist lead at Mastercard, I spend up to half of my day in meetings, both with my smaller tech team to support other data scientists to solve technical issues they are facing and with the wider Mastercard product teams to report or catch up on projects and plan for future projects.

The rest of the time is mainly for technical works such as architecting, designing, coding and problem solving.

Currently, machine learning and big data are the two AI skills that I am using the most in daily projects, in addition to SQL which is needed for analytics.

However, the most important elements that I am developing in my current role are management and communication skills. I try to pull key learnings that I see from experienced managers and apply them to manage my smaller tech team.

The hardest part is the data itself. Getting the correct data, and clean data, for the right project is important. This work cannot be solved alone, it needs support from the whole team.

An area of growing importance in the field of AI and data analytics is that of model explainability this concept relates to how we can explain or interpret, in human terms, how a model is reaching its prediction or decision. This is an important part of Mastercards commitment to ethical AI, and means that we need to consider how we can achieve an appropriate level of explainability when we embark on developing a new model.

If you have free time, try to read technical books. Its incredibly productive to use time to read while learning and improving your technical expertise.

Im a big believer in trying to automate your work as much as possible. Automating repetitive tasks will free you from any mistakes caused by human error while also saving you time and energy that can be used to do other tasks. In my work, I find that I can often use existing tools such as Jenkins and Docker to automate the CI pipeline for running unit tests, or Ill develop a simple tool in Python or some other scripting language to do it myself.

Tools like Jira, ALM and Confluence are very efficient for daily work collaboration. While Jira and ALM help to manage the progress of projects efficiently, Confluence is a place where you can store documents and share knowledge with your colleagues.

These tools are very useful for advancing the efficiency of our collaboration besides other common communication tools such as email, instant messaging and coffee tables.

In the past, data scientists were often familiar with data science skills but not engineering skills. However, full stack data scientists nowadays should train themselves for the engineering skills such as Docker containerisation, Kubernetes, AWS cloud and CICD (machine learning/dev-ops).

If you have free time, try to read technical books

It is also recommended that data scientists should practice standard coding styles (such as using Git professionally, following PEP8 rules, writing unit tests, etc.)

Joining an AI project and knowing that I am creating a product that is useful for our community makes me happy. I used to contribute to an open-source AI project which has thousands of users, so I feel like I am helping people to make the world a better place.

I also push myself to continuously innovate and I work hard to secure more and more patents for the company for the newest technology.

The most important thing for success in AI work is to practice as much as possible on multiple types of data. For example, joining Kaggle competitions is a good starting point for junior data scientists, where they have multiple types of free data to practise with. Furthermore, success in these competitions also gives people more confidence and advances their career. In addition, practical tutorials of deep learning frameworks such as Tensorflow and Pytorch are also helpful to practice in AI.

Dont forget to always be innovative writing patents will help refresh your mind and keep you motivated.

In addition, I recommend that people working in AI should read. I like to read good technical books and often read them twice. This habit helps extend your knowledge efficiently to help your job and your career in the long term. Pattern Recognition by Christopher Bishop is an enjoyable book that helps to master machine learning and I must have read this at least three times. Deep Learning by Ian Goodfellow, Aaron Courville and Yoshua Bengio is also a book that you cannot miss when working in AI.

10 things you need to know direct to your inbox every weekday. Sign up for the Daily Brief, Silicon Republics digest of essential sci-tech news.

See more here:

The hardest part about working in AI and analytics 'is the data itself' - SiliconRepublic.com

This article is part of a Nature series examining data on ethnic or racial diversity in science in different countries. See also: How UK science is failing Black researchers and How Indias caste system limits diversity in science.

In November 2021, Isolde Karles university in Germany appointed her to a newly created high-ranking role, which signalled the attention it was paying to diversity. As vice-rector for diversity, inclusion and talent development at the Ruhr University Bochum, Karles job involves helping students from minority groups including those discriminated against because of their skin colour to participate fully in science.

There is racism in society, and we universities are not free of it, says Karle, who is also a theologian and the universitys pastor.

Survey data support Karles concerns. This January, for instance, Germanys government issued its first annual report on racism. It cites, among other data, a large telephone survey that found more than one-fifth of people in the country had personally experienced racism (see go.nature.com/3zqsmvt; in German). And at universities, a Nature worldwide survey of graduate students last year found that, in Germany, 6% had experienced racial discrimination or harassment, but that this rose to 19% among those who identified as being in a racial or ethnic minority; findings similar to the surveys global results (see German graduate experiences).

Source: Nature graduate survey 2022 (Nature https://doi.org/j4cn; 2022)

Karle has talked to some researchers from minority ethnic groups at her university about their experiences. But she cant clearly track racial or ethnic diversity among the scholars at her institution or how it might be changing, because she has no data. That is the norm for employers all over Germany, where there is general unease at the idea of collecting information on ethnicity even though doing so is not legally prohibited as well as a feeling that it is not a priority issue, according to many Nature talked to.

Such concerns are often explained with reference to the countrys history and the horrific example of how the Nazis used census data to organize genocide. As a result, sensitivity over data protection and privacy is high: surveying people in Germany about their ethnicity has a strong negative connotation.

Karle expresses these reservations, too; asking employees about their race or heritage can itself be discrimination, she explains. We want to make the differences that lead to discrimination disappear, but by asking about them, we would focus on them, she says. This is part of the paradox.

The result in Germany is that, unlike in the United Kingdom, United States and several other countries, there are no census data on racial or ethnic diversity, and neither research funders nor universities collect this information. A spokesperson for the Bonn-based German Rectors Conference, an association of 269 universities in Germany, told Nature that there was no recording of personal characteristics that could identify minority ethnic groups for understandable and historical reasons, and because of data protection.

It is a similar situation in some other European countries: three research-funding agencies in France, Sweden and the Netherlands, for instance, told Nature that they dont collect diversity data on race or ethnicity, generally citing data-protection concerns.

Of course, this makes it difficult to quantitatively measure success on ethnic diversity issues, says a spokesperson for Germanys Max Planck Society (MPS) in Munich, a large public association of research institutes.

But this cultural unease is starting to shift. Some researchers in countries including Germany are pushing to gather data on ethnicity and race in academia, arguing that the advantages of quantitative data collection outweigh the concerns. In July 2022, Germanys largest public research funder, the DFG, issued guidelines that put a greater emphasis on ethnic diversity in data gathering.

For now, there are only snapshots of information. In 2020, an association of PhD students at the MPS, called Max Planck PhDnet, organized its own survey, asking more than 2,000 doctoral students about their citizenship and ethnicity (P.-G Majev et al. 2020 Survey Report https://doi.org/j39n; Max Planck PhDnet, 2021). It reported that 71% of the students who provided an ethnicity described themselves as of European descent, 10% as of East Asian or Southeast Asian descent, 7% as of South Asian descent and 0.7% (16 people) as of African descent. One-third of people describing themselves as non-European said they had felt discriminated against at work, compared with 18% of respondents identifying as European (see Representation among German doctoral students).

Source: P.-G Majev et al. https://doi.org/j39n (Max Planck PhDnet, 2021)

These findings are comparable to those in Natures 2022 survey: almost one-quarter of German respondents identified as being in an ethnic or racial minority and, of those, more than one-third said they had experienced discrimination or harassment (not always racial); compared with 17% of the other German respondents.

Larger surveys have tended to collect data using a coarse proxy called migration background. This typically refers to a person with at least one parent who was not born a German citizen which doesnt necessarily map well on to minority racial or ethnic status. In 2019, the German Centre for Higher Education Research and Science Studies (DZHW), a research institute in Hanover, launched a longitudinal study called the National Academics Panel Study that aims to examine the demographics and career paths of doctoral researchers in Germany. It has started to collect data on factors that include migration background and country of birth, but not ethnicity (see go.nature.com/3kt68jv; in German).

The DZHWs publicly available data define migration background as people who were born outside Germany. It found that this group accounted for 24% of some 14,000 doctoral researchers surveyed in 201920. (Data using the wider definition of migration background that is, including PhD students who have a parent born outside Germany are available to researchers on request, the DZHW says). In a separate study of around 18,000 German students (including undergraduates) last year, the DZHW found that 6% of all students said they had been discriminated against because of their migration background (see go.nature.com/3kgcdj7; in German).

How Indias caste system limits diversity in science in six charts

And last September, the German Education Union (GEW), which represents researchers and teachers, analysed data from multiple sources to suggest that 2% of German schoolchildren with a migration background might go on to gain a doctorate, whereas those with two German-born parents had double the chance, at 4% (see go.nature.com/3tzwcuj; in German). (The union added that the main differences occurred in transitions from primary school to high school, and from high school to bachelors degrees, rather than later in the academic pipeline.)

Such reports leave many questions about diversity in German science unanswered. But the picture might be changing. The Berlin non-governmental organization Citizens for Europe (CFE), for instance, aims to survey the full diversity of people in Germany including the interplay of race, culture and social class, as well as gender. In 2021, CFE and another Berlin-based non-governmental organization, Each One Teach One (EOTO), released the results of Afrozensus, a survey of more than 6,000 Black, African and Afro-Diasporic people in Germany, which aimed to raise awareness about the experiences of Black people in the country (see go.nature.com/3nptj9m; in German). It did not specifically analyse academic professions, but more than 4% of respondents had a PhD.

In all, only 15.3% of Afrozensus respondents said they had never been discriminated against in a professional context. The lack of data on race and discrimination is part of a vicious circle, says Daniel Gyamerah, who worked on the survey while at CFE. He is now head of a new non-profit think tank, called zedela (Centre for Data Driven Empowerment, Leadership and Advocacy), and is also EOTOs chairperson. Without data, it is hard to grasp the extent of discrimination problems, Gyamerah says, which then makes it harder to convince organizations and policymakers that they should track the issue. Reservations over data-protection laws are often used as an excuse to avoid engaging with the topic, he says.

Sociologist Lucienne Wagner, a CFE researcher who is pursuing a PhD at the University of Koblenz-Landau on the discourses of diversity at German universities, sees a further challenge for raising awareness about racial discrimination and collecting data on diversity at universities. With scant resources, people focused on gender equality can end up competing against rather than working with those exploring other dimensions of discrimination. But efforts in the field mean that attitudes in academia are changing, she adds.

The DFGs diversity guidelines, issued in July, could mark a significant shift. Transparency of equality requires the collection and publication of continuous, differentiated data at all academic career levels, in particular concerning the participation of women and men. The same applies to other dimensions of diversity as far as legally permissible, the guidelines say.

Asked for more details, the DFG told Nature that it doesnt have any specific rules on how to collect data on cultural or ethnic representation besides recommending anonymous surveys that pay careful attention to data protection. The funder says that it is up to universities to decide how to conduct such surveys and what data to collect, and that it doesnt have an overview of these activities. Universities and research institutions can and should already take measures to promote diversity in view of the evidence that is already available, a spokesperson said, referring to surveys by the DZHW and others. But the DFG also said that it is not aware of any universities already implementing such measures.

Last September, the German Rectors Conference launched a federally funded initiative called Diversity at German Universities. Details are limited, but the aim is to fund some universities to examine what hurdles and barriers need to be overcome to enable more diversity and inclusivity, with results due to be presented in 2024; monitoring diversity is also a goal.

How UK science is failing Black researchers in nine stark charts

For Joshua Kwesi Aikins, a political scientist and human-rights activist at the University of Kassel, who worked on the Afrozensus for CFE and is joining zedela, the time to collect information on race and on multiple forms of discrimination has already come. However, he adds, such data must be collected only with the full understanding and participation of the groups who are being asked for the information, so that the exercise is done in good faith.

History has led Germany to be rightfully cautious about racial classification. But German institutions need to act, he says. A strikingly male-dominated academe perpetuates racist, sexist and classist dynamics that add to structural hurdles on the way to what remains for too many a precarious academic career.

Aikins and Wagner are part of a three-year research network funded by the DFG and running until 2025, which aims to define new categories for data collection. This could include self-reported ethnicity, among other facets of identity. The group, coordinated by Anne-Kathrin Will, who studies migration at the Humboldt University of Berlin, and Linda Supik, a sociologist at Leibniz University in Hanover, is working with non-governmental and civil-society organizations.

Whether these efforts will change how German employers or universities collect race or ethnicity data remains to be seen. The DFG says that, in future, researchers wanting funding will need to address how they are promoting diversity, and that it plans to develop individual measures for this over a period of approximately five years. It also notes that it is developing a survey to understand needs and expectations related to its promotion of diversity, but says it cannot comment further.

At the Ruhr University Bochum, meanwhile, Karle says work is being done to improve the understanding of concerns over discrimination. Last July, for instance, Hans Alves, a psychologist at the university who specializes in social cognition, ran a survey of more than 4,000 students asking about their gender, migration status, religion and experiences of various kinds of discrimination. It did not ask about race or ethnicity. The results have not yet been published, but Alves told Nature the survey showed that 10% of students with a parent born outside Germany said they had been the target of discrimination because they were perceived as foreign or non-white. Karle says the university is now thinking about conducting a similar project across all of its academic staff.

Originally posted here:

Diversity in German science: researchers push for missing ethnicity ... - Nature.com

Venture capital has traditionally been an industry that revolves around relationships. VCs invest in a startups idea but their conviction stems from the folks behind it. This largely makes sense because investing in a startup also usually entails entering a years-long relationship.

But backing companies based on the allure of the founder hasnt always worked out. Indeed, it often gets investors tied up in companies destined to collapse for one reason or another. And depending on warm intros or networks also limits the amount of startups an investor considers, which further alienates founders who dont have the same networks or hail from nontraditional backgrounds.

An increasing number of venture firms think the solution to cutting through the noise is by incorporating data science into their deal sourcing process. This wouldnt be a crazy idea per se, as investors from other asset classes such as institutional investors, hedge funds and public market traders already embrace data-driven investing, but thus far venture has largely sat out of the conversation.

A few venture firms, such as Correlation Ventures, SignalFire and Rocketship.vc, have long taken this approach, but this number looks likely to grow.

This week, Austin, Texas-based VC outlet Ensemble announced that it closed a $100 million debut fund to invest in early-stage startups using a data-driven approach that sorts and tracks companies based on the quality and depth of their entire team.

Ensembles co-founder and managing partner, Collin West an alum of Correlation Ventures told TechCrunch that the firm wants to back companies that have the strongest team, but that would be too difficult to track without using data science to pare the list down.

Using software, we can track all of the people at all of the startups, and that ends up being a whole lot more information than any human brain can handle, and especially any venture firm, West said. We effectively sort the industry by team quality in a very objective way knowing which companies to focus on, and spend a lot more time on fewer companies.

Continued here:

Slowly but surely, data is helping VCs look beyond networks for ... - TechCrunch

Manchester's Gay Village. (Photo by Christopher Furlong/Getty Images)Getty Images

Active Travel England has commissioned the Alan Turing Institute to create software and data science techniques to support local authority delivery of walking and cycling schemes. The institute, based out of the British Library in London, is the UKs national data science and artificial intelligence institute.

The collaboration will run for two years at a total cost of $250,000 and enable the development of new functionality in the Active Travel Infrastructure Platform (ATIP), which helps councils to map out proposed schemes and see the impact they could have locally.

These new tools will be paired with existing data sources, such as OpenStreetMap, to create solutions that will help build the evidence needed to meet the national governments stated objectives on active travel, including for 50% of short trips in urban areas to be made by walking, wheeling and cycling by 2030.

The new software engineering and data science techniques will complement data collection and analysis work done by Active Travel Englands head of data Dr. Robin Lovelace.

The partnership with Alan Turing Institute is hugely important, Lovelace told Forbes.com.

Transport models and datasets represent leverage points in the transport planning system, he stressed.

The lack of data and robust analysis of active modes has led to them not being taken seriously. New datasets can ensure that investment goes where its most needed.

Active Travel Minister Jesse Norman said the partnership will enable local councils to draw on the latest technology and maximize active travel's environmental, economic and health benefits.

Meanwhile, the government he represents last month revealed swingeing cuts to Englands active travel budget.

According to the Walking and Cycling Alliance (WCA), a body made up of cycling and walking organizations including British Cycling, Living Streets, Ramblers and Sustrans, the cuts means a two-thirds cut to promised capital investment in walking and cycling.

It is heartbreaking to see vital active travel budgets wiped away in England, at the exact time when they are most essential to U.K. economic, social and environmental prospects, said a WCA statement.

It is incredibly disappointing that the active travel budget has seen such extensive cuts at a time where we need to really make progress on decarbonisation and when people need cheap transport choices, said a joint statement by Conservative MP Selaine Saxby and Labour MP Ruth Cadbury, co-chairs of the All Party Parliamentary Group on Cycling and Walking.

They added: We understand that there are pressures on the public purse but active travel schemes frequently have much higher benefit:cost ratios than road building schemes, many of which are still going ahead despite falling value for money for taxpayers.

I was Press Gazette's Transport Journalist of the Year, 2018. I'm also an historian my most recent books include "Roads Were Not Built for Cars" and "Bike Boom", both published by Island Press, Washington, D.C.

See the original post here:

Active Travel England Partners With Alan Turing Institute To Leverage Data Into Investment - Forbes

Data science is an expeditiously growing field that involves analyzing, interpreting, and making predictions based on large sets of data. As the field continues to expand, these programs have become progressively popular as a way for individuals to validate their skills and knowledge. In this article, we will understand various types of data science programs available, the benefits of obtaining a certification, and factors to consider when selecting a program.

Data science programs come in different types and are offered by various providers. The most common types of certification programs are vendor-specific, industry-specific, and vendor-neutral.

Vendor-specific certification programs are offered by technology companies, such as Microsoft and Amazon, and are designed to validate an individuals proficiency in their specific technologies. Industry-specific certification programs, on the other hand, are tailored to a particular industry, such as healthcare or finance. Vendor-neutral certification programs, such as those offered by the Data Science Association and the Institute of Electrical and Electronics Engineers (IEEE), are not tied to a specific technology or industry. Ed-tech platforms like Great Learning also offer well-curated programs in association with world-class universities.

The requirements for certification programs vary depending on the provider and the type of program. Some programs require passing an exam, while others require completing a course or submitting a project.

There are several benefits to obtaining a data science certification. Here are some of the most significant benefits:

Certification programs provide recognition and credibility to individuals, validating their skills and knowledge in the field. This recognition can be particularly valuable for job seekers, as it demonstrates to potential employers that they have the necessary qualifications for the job.

Many certification programs offer additional resources, such as online forums, workshops, and study materials, which can help you stay up-to-date with the latest trends and best practices in the field.

A data science certification provides tangible evidence of your knowledge and expertise in the field. Employers and colleagues can see that you have invested time and effort into gaining a deep understanding of the subject matter.

Data science certifications can also open up new career opportunities and advancement paths. Individuals with certifications are often preferred by employers, which can lead to promotions, salary increases, and other benefits.

Certification programs can also help individuals enhance their skills and acquire new ones. The coursework and exams required for certification can help individuals develop new skills and deepen their understanding of the field.

Certification programs can also provide networking opportunities. Individuals in certification programs can connect with other professionals in the field, allowing them to expand their network and potentially find new job opportunities.

Learning data science can lead to higher salary potential. According to a study, certified professionals in the IT field can earn up to 12% more than their non-certified peers.

When selecting a data science certification program, its essential to consider several factors, including:

When choosing a certification program, it is important to consider all of these factors in order to select the best program for your needs. Research the provider, determine the requirements, and consider the cost to ensure you are making the right decision for your future.

In conclusion, pursuing post-graduation in data science can provide significant benefits to individuals looking to enhance their skills. They can provide the opportunity to develop data science expertise and apply it to a specific industry. Furthermore, they offer the chance to learn the latest technologies and techniques, as well as become a certified data scientist. Certificate programs provide a great way to gain recognition and credibility in the workplace, as well as to further your career. Therefore, if you are looking to break into the data science field or simply want to expand your existing knowledge, looking into data science programs could be the perfect choice for you.

Read more here:

Data Science Certification Programs and Their Benefits - IndiaCSR