Native DNA is so sought after that people are looking for proxy data, and one of the big proxy data is the microbiome Mr. Yracheta said. If youre a Native person, you have to consider all these variables if you want to protect your people and your culture.

In a presentation at the conference, Joslynn Lee, a member of the Navajo, Laguna Pueblo and Acoma Pueblo Nations and a biochemist at Fort Lewis College in Durango, Colo., spoke about her experience tracking the changes in microbial communities in rivers that experienced a mine wastewater spill in Silverton, Colo. Dr. Lee also offered practical tips on how to plan a microbiome analysis, from collecting a sample to processing it.

In a data-science career panel, Rebecca Pollet, a biochemist and a member of the Cherokee Nation, noted how many mainstream pharmaceutical drugs were developed based on the traditional knowledge and plant medicine of Native people. The anti-malarial drug quinine, for example, was developed from the bark of a species of Cinchona trees, which the Quechua people historically used as medicine. Dr. Pollet, who studies the effects of pharmaceutical drugs and traditional food on the gut microbiome, asked: How do we honor that traditional knowledge and make up for whats been covered up?

One participant, the Lakota elder Les Ducheneaux, added that he believed that medicine derived from traditional knowledge wrongly removed the prayers and rituals that would traditionally accompany the treatment, rendering the medicine less effective. You constantly have to weigh the scientific part of medicine with the cultural and spiritual part of what youre doing, he said.

Over the course of the IndigiData conference, participants also discussed ways to take charge of their own data to serve their communities.

Mason Grimshaw, a data scientist and a board member of Indigenous in A.I., talked about his research with language data on the International Wakashan A.I. Consortium. The consortium, led by an engineer, Michael Running Wolf, is developing an automatic speech recognition A.I. for Wakashan languages, a family of endangered languages spoken among several First Nations communities. The researchers believe automatic speech recognition models can preserve fluency in Wakashan languages and revitalize their use by future generations.

Original post:

Training the Next Generation of Indigenous Data Scientists - The New York Times

Sven Melsom Ziegler ex Clarkson Platou will head and build Maritime Optimas analysis/data science unit. Sven is a well-known and experienced Shipping/Offshore Market Analyst.

I am looking forward to working with the team and the data in Maritime Optima, Mr Melsom Ziegler says. I was introduced to the team and their data platform during the spring. It is rare to find a startup investing so much in maritime data quality and seeing such a competent team with so much passion and willingness to improve data quality continuously. Since the industry lacks clear definitions of most data, someone must start doing this thoroughly because machines need clear definitions to create value for humans. Otherwise, the gains from digitalization and automation will be hard to obtain. Here we have a unique data platform that enables state-of-the-art shipping/offshore models and analysis.

Sven Melsom Ziegler started working with RS Platou, where he stayed for 21 years. Sven was born in Cape Town and raised in Larvik and Athens. After completing his education at Strathclyde University of Glasgow and CASS in London, he later settled down in Oslo. He comes from Forum Market Services, where he has been working with mainly oil market/bunker analysis and quantitative strategies.

We are very proud to have Sven on board, says the founder and CEO Kristin Omholt-Jensen. We have kept a very low profile so far. During the Covid months, we have spent time keeping the team well, together and motivated, investing in and collecting data, and defining data templates, testing and scaling. It is super important for us to develop the product together with our users, so since our R&D partners left their offices and went to their home offices while we needed active feedback from real users, we launched a freemium application last autumn. Since the launch, we have been growing very quickly. Today we have close to 9000 active registered users, and we pick up AIS data every minute for more than 65 000 vessels from more than 600 satellites and terrestrial senders. We also know we manage to do cost-effective scaling. We have a very exciting road-map and will develop new products / features continuously (based on feedback from the users) and it is perfect timing to have Sven as a part of our team.

Maritime Optima is set up to develop and distribute user-friendly and flexible maritime saas software across platforms helping professionals in the maritime industry save time, work more efficiently, make better decisions and maybe have more fun. Software should be easy to understand and looked upon as a partner and not something you have to use and hate. We believe that colleagues should be working in teams, but you can also work in your one-man-team if you want. We think professionals will dare to share more and more, but they will not want to share everything with everyone. So we have made it super-easy to share public or keep private, and we have therefore included a user log showing the users and teams activity, so it is easy to find later.

The maritime office software industry is young, and there are many startups. To change the way the industry works, their routines, and how things have been done for years will take time, but we are prepared and willing to show that we are here to stay. We continuously launch new features based on user feedback, invest time to improve the data quality and increase the number of users. The maritime software industry is still young and will be consolidated during the next few years, and we want to play an active role in that consolidation, says Omholt-Jensen.

Start building your own maritime office by registering a free account in Maritime Optima: http://www.app.maritimeoptima.comSource: Maritime Optima

Read more from the original source:

New High Profile Addition to Maritime Optimas Team, Set to Build the Companys Analysis/Data Science Unit - Hellenic Shipping News Worldwide

Last autumn, following a difficult six months of redundancies and furloughs that hit the arts and culture sector particularly hard, a governmental campaign encouraging people to retrain to work in the tech industry was met with heavy criticism from the public. The ad showed a young dancer named Fatima tying up her ballet shoes, likely in preparation for rehearsal, with the message Fatimas next job could be in cyber. (she just doesnt know it yet) overlaid on the left of the image. In the end, the campaign became so unpopular (not to mention thoroughly ridiculed) that the campaign was ultimately scrapped and the government forced to apologise.

Although it was probably not the main intention by design, the ad became a symbol of the way the arts and culture sectors are seen as frivolous, held in less respect than the supposedly more responsible and important STEM subjects. Last year, arts and design, which could well have been the course studied by the person hired by the government to create the infamous campaign, was ranked ninth out of the ten most popular university subjects, according to QS World University Rankings by Subject. Computer science and information systems and engineering and technology topped the list, while humanities subjects such as history, languages, literature, and philosophy, were nowhere to be seen. This shouldn't come as a surprise: Many young people considering studying Arts & Humanities are advised not to pursue this path, as its stereotypically seen as a fast track to unemployment or redundancy, as seen during the pandemic. In fact, if you imagine a hypothetical usefulness spectrum dictated by the economy, humanities are going to be on the opposite end from subjects such as maths, computer science, and chemistry.

Long-time NASA engineer Peter Scott uses a metaphor of driving a car to illustrate the divide between those who are in the tech industry and those who arent.

The fact is that it's a matter of perspective, and this perspective came to me after driving my children everywhere. They have to sit in the back seat all the time and when one of them gets old enough, and they get to ride in the front seat once in a while, it's like: Oh, you can see so much more here.

In a world that is rapidly progressing with new technologies, being outside of STEM is a bit like being driven around in a car while being forced to sit in the back seat.

There are insiders and outsiders in every industry, but the tech industry is the one that's doing the most to reshape where we're going, he says. For everyone else, it's like: We don't know where we're going. [The driver] seems to think you know where you're going, but you're not giving us a good enough picture of it.

After three decades at NASA's Jet Propulsion Laboratory (JPL), Scott decided to embark on a more human-facing career of public speaking, which he describes as scarier and less likely to happen than jumping out of a plane. However, he also notes that he managed to achieve both. Nowadays, he blends being a business coach and successful TEDx speaker with contracting for NASA.

I'm balancing both of these worlds because I want and need to be able to see both sides at the same time. Engineers and scientists, we tend to get locked into a certain view of the world that's driven by the equations and the principles that we know and the natural laws that explain everything. And you can't say that it doesn't, because that's the principle of science, he says.

However, while science may be the vanguard of change, if taken in isolation it leaves out a whole perspective that's driven by poetry, emotion, and artistic values, notes Scott.

Because they're not quantifiable and measurable, they get not so much disdained as just ignored by scientists and tech people. It doesn't get you where you need to go as a tech person, so you can't afford to spend time on that. So these worlds are like C P Snows Two Cultures theyre growing, if anything, further apart, he warns.

The gap between the two separate cultures of humanities and STEM is especially visible in the evolving technology of artificial intelligence, which is becoming more present in everyday life in our phones, workplaces, and even supermarkets.

Our journey with AI especially is one that requires a common understanding, says Scott. We can't advance this technological agenda that upends everyone's life in a largely, and hopefully, positive way without understanding both sides, without finding the bridge between those two cultures.

Although AI is treated as inherently technological, last years events have proven that its also a major ethical issue. Nevertheless, despite University of Oxford physicist David Deutsch predicting in 2012 that philosophy will be the key that unlocks artificial intelligence, the New College of the Humanities (NCH) is so far the only university in the UK to offer a joint degree in philosophy and artificial intelligence. Dr Brian Ball, who is the head of NCHs philosophy faculty and an associate professor, tells IT Pro that the degree was launched after the school partnered with the Boston-based Northeastern University.

We are quite proud of our MA in Philosophy and Artificial Intelligence, and some of our related degrees, such as our MSc AI with a Human Face, and our various bachelor's degrees with humanities majors and data science minors, he says. They are prompted in part by our joining the global network of Northeastern University, where interdisciplinarity and the cultivation of human, data, and technological literacies are central to higher education, and partly by the intrinsic merits of studying these subjects together.

The IT Pro Podcast: Soft skills vs STEM skills

The UKs level of soft skills is dropping - so what can we do about it?

According to Ball, AI can benefit from at least two of these merits, with philosophy able to provide the technology with ethicality as well as explainability. This is particularly crucial at a time when facial recognition is increasingly under fire for being prone to unethical usage.

Artificial intelligence isnt the only field where philosophy might be useful, however. Over the last five years, Exasol chief data and analytics officer Peter Jackson recruited plenty of data scientists, but not all of them come from a traditional IT or data science background. In fact, he says the best data scientist to ever be a part of one of his teams, who was creative, curious, and could turn insights into compelling arguments, didn't hold a degree in computing or data science, but in philosophy. Jackson says that, when recruiting, he doesnt only look at candidates technical skills and their ability to understand data, but also their storytelling skills. According to him, this specific ability can be found in somebody who's done English literature, who is very good at writing poems and stories, and building a coherent argument.

I need them to be able to interpret the output of that piece of work, either to me, or the rest of the team, or to stakeholders, he tells IT Pro. If they can only go so far, and they have to hand [it] over to somebody else to tell the story, you can get a disconnect. The person who's telling the story might not be able to answer some of the technical questions that may arise: What training set did you use? or Where did that data come from?. So I try to recruit data scientists who are able to at least tell the first part of the story of their work.

However, Jackson notes that finding people with both data science and storytelling skills is very hard.

Sometimes, because of particular skills that you need from the data science point of view, you quite often compromise on that. And that's where you do need the professional storytellers, professional writers who can support it, he adds.

Asked about his thoughts on the split between STEM sciences and humanities, Jackson says that he doesnt see it as a dichotomy.

I dont think there should be a divide. I think as a society, as an economy, we need smart, educated people and I think that is the priority.

Owning your own access security

The key to building strong cloud security and avoiding the risk of vendor lock-in

Developing a dynamic infrastructure

How to implement holistic changes to support distributed workers

How upgraded server and storage platforms support digital transformation

New Dell EMC PowerStore delivers high-end enterprise storage features at midrange price point

How to maximise the value of your data and apps with IaaS

Free yourself from infrastructure complexity

See the original post:

Humanities versus STEM: The forced dichotomy where no one wins - IT PRO

Linear regression, while a useful tool, has significant limits. As its name implies, it cant easily match any data set that is non-linear. It can only be used to make predictions that fit within the range of the training data set. And, most importantly for our purposes, linear regression can only be fit to data sets with a single dependent variable and a single independent variable.

This is where multiple regression comes in. While it cant overcome all of linear regressions weaknesses, its specifically designed to create regressions on models with a single dependent variable and multiple independent variables.

Multiple regression is an extension of linear regression models that allow predictions of systems with multiple independent variables.

To start, lets look at the general form of the equation for linear regression:

y = B * x + A

Here, y is the dependent variable, x is the independent variable, and A and B are coefficients dictating the equation. The difference between the equation for linear regression and the equation for multiple regression is that the equation for multiple regression must be able to handle several inputs, instead of only the single input of linear regression. To account for this change, the equation for multiple regression looks like this:

y = B_1 * x_1 + B_2 * x_2 + + B_n * x_n + A

In this equation, the subscripts denote the different independent variables. For example, x_1 is the value of the first independent variable, x_2 is the value of the second independent variable, and so on. It keeps going as we add more independent variables until we finally add the last independent variable, x_n, to the equation. (Note that this model allows you to have any number, n, independent variables and more terms are added as needed.) The B coefficients employ the same subscripts, indicating they are the coefficients linked to each independent variable. A, as before, is simply a constant stating the value of the dependent variable, y, when all of the independent variables, the xs, are zero.

As an example, imagine that youre a traffic planner in your city and need to estimate the average commute time of drivers going from the east side of the city to the west. You dont know how long it takes on average, but you do know that it will depend on a number of factors like the distance driven, the number of stoplights on the route, and the number of other cars on the road. In that case you could create a linear multiple regression equation like the following:

y = B_1 * Distance + B_2 * Stoplights + B_3 * Cars + A

Here y is the average commute time, Distance is the distance between the starting and ending destinations, Stoplights is the number of stoplights on the route, and A is a constant representing other time consumers (e.g. putting on your seat belt, starting the car, maybe stopping at a coffee shop).

Now that you have your commute time prediction model, you need to fit your model to your training data set to minimize errors.

Similarly to how we minimize the sum of squared errors to find B in linear regression, we minimize the sum of squared errors to find all of the B terms in multiple regression. The difference here is that since there are multiple terms, and an unspecified number of terms until you create the model, there isnt a simple algebraic solution to find A and B. This means we need to use stochastic gradient descent. You can find a good description of stochastic gradient descent in Data Science from Scratch by Joel Gros or use tools in the Python scikit-learn package. Fortunately, we can still present the equations needed to implement this solution before reading about the details.

The first step is summing the squared errors on each point. This takes the form:

Error_Point = (Actual Prediction)

In this instance, Error is the error in the model when predicting a persons commute time, Actual is the actual value (or that persons actual commute time), and Prediction is the value predicted by the model (or that persons commute time predicted by the model). Actual Prediction yields the error for a point, then squaring it yields the squared error for a point. Remember that squaring the error is important because some errors will be positive while others will be negative and if not squared these errors will cancel each other out making the total error of the model look far smaller than it really is.

To find the error in the model, the error from each point must be summed across the entire data set. This essentially means that you use the model to predict the commute time for each data point that you have, subtract that value from the actual commute time in the data point to find the error, square that error, then sum all of the squared errors together. In other words, the error of the model is:

Error_Model = sum(Actual_i Prediction_i)

Here i is an index iterating through all points in the data set.

Once the error function is determined, you need to put the model and error function through a stochastic gradient descent algorithm to minimize the error. The stochastic gradient descent algorithm will do this by minimizing the B terms in the equation.

Once youve fit the model to your training data, the next step is to ensure that the model fits your full data set well.

To make sure your model fits the data use the same r value that you use for linear regression. The r value (also called the coefficient of determination) states the portion of change in the data set predicted by the model. The value will range from 0 to 1, with 0 stating that the model has no ability to predict the result and 1 stating that the model predicts the result perfectly. You should expect the r value of any model you create to be between those two values. If it isnt, retrace your steps because youve made a mistake somewhere.

You can calculate the coefficient of determination for a model using the following equations:

r = 1 (Sum of squared errors) / (Total sum of squares)

(Total sum of squares) = Sum(y_i mean(y))

(Sum of squared errors) = sum((Actual_i Prediction_i))

Heres where testing the fit of a multiple regression model gets complicated. Adding more terms to the multiple regression inherently improves the fit. Additional terms give the model more flexibility and new coefficients that can be tweaked to create a better fit. Additional terms will always yield a better fit to the training data whether the new term adds value to the model or not. Adding new variables which do not realistically have an impact on the dependent variable will yield a better fit to the training data, while creating an erroneous term in the model. An example of this would be adding a term describing the position of Saturn in the night sky to the driving time model. The regression equations will create a coefficient for that term, and it will cause the model to more closely fit the data set, but we all know that Saturns location doesnt impact commute times. The Saturn location term will add noise to future predictions, leading to less accurate estimates of commute times even though it made the model more closely fit the training data set. This issue is referred to as overfitting the model.

Additional terms will always improve the model whether the new term adds significant value to the model or not.

This fact has important implications when developing multiple regression models. Yes, you could keep adding more terms to the equation until you either get a perfect match or run out variables to add. But then youd end up with a very large, complex model thats full of terms which arent actually relevant to the case youre predicting.

One way to determine which parameters are most important is to calculate the standard error of each coefficient. The standard error states how confident the model is about each coefficient, with larger values indicating that the model is less sure of that parameter. We can intuit this even without seeing the underlying equations. If the error associated with a term is typically high, that implies the term is not having a very strong impact on matching the model to the data set.

Calculating the standard error is an involved statistical process, and cant be succinctly described in a short article. Fortunately there are Python packages available that you can use to do it for you. The question has been asked and answered on StackOverflow at least once. Those tools should get you started.

After calculating the standard error of each coefficient, you can use the results to identify which coefficients are highest and which are lowest. Since high values indicate that those terms add less predictive value to the model, you can know those terms are the least important to keep. At this point you can start choosing which terms in the model can be removed to reduce the number of terms in the equation without dramatically reducing the predictive power of the model.

Another method is to use a technique called regularization. Regularization works by adding a new term to the error calculation that is based on the number of terms in the multiple regression equation. More terms in the equation will inherently lead to a higher regularization error, while fewer terms inherently lead to a lower regularization error. Additionally, the penalty for adding terms in the regularization equation can be increased or decreased as desired. Increasing the penalty will also lead to a higher regularization error, while decreasing it will lead to a lower regularization error.

With a regularization term added to the error equation, minimizing the error means not just minimizing the error in the model but also minimizing the number of terms in the equation. This will inherently lead to a model with a worse fit to the training data, but will also inherently lead to a model with fewer terms in the equation. Higher penalty/term values in the regularization error create more pressure on the model to have fewer terms.

Read More From Our ExpertsWhat is Linear Regression? Explaining Concepts and Applications With Tensorflow 2.0

The model youve created is not just an equation with a bunch of numbers in it. Each one of the coefficients you derived states the impact an independent variable has on the dependent variable assuming all others are held equal. For instance, our commute time example says the average commute will take B_2 minutes longer for each stoplight in a persons commute path. If the model development process returns 2.32 for B_2, that means each stoplight in a persons path adds 2.32 minutes to the drive.

This is another reason its important to keep the number of terms in the equation low. As we add more terms it gets harder to keep track of the physical significance (and justify the presence) of each term. Anybody counting on the commute time predicting model would accept a term for commute distance but will be less understanding of a term for the location of Saturn in the night sky.

Note that this model doesnt say anything about how parameters might affect each other. In looking at the equation, theres no way that it could. The different coefficients are all connected to only a single physical parameter. If you believe two terms are related, you could create a new term based on the combination of those two. For instance, the number of stoplights on the commute could be a function of the distance of the commute. A potential equation for that could be:

Stoplights = C_1 * Distance + D

In this case, C_1 and D are regression coefficients similar to B and A in the commute distance regression equation. This term for stoplights could then be substituted into the commute distance regression equation, enabling the model to capture this relationship.

Another possible modification includes adding non-linear inputs. The multiple regression model itself is only capable of being linear, which is a limitation. You can however create non-linear terms in the model. For instance, say that one stoplight backing up can prevent traffic from passing through a prior stoplight. This could lead to an exponential impact from stoplights on the commute time. You could create a new term to capture this, and modify your commute distance algorithm accordingly. That would look something like:

Stoplights_Squared = Stoplights

y = B_1 * Distance + B_2 * Stoplights + B_3 * Cars + B_4 * Stoplights_Squared + C

These two equations combine to create a linear regression term for your non linear Stoplights_Squared input.

Multiple regression is an extension of linear regression models that allow predictions of systems with multiple independent variables. We do this by adding more terms to the linear regression equation, with each term representing the impact of a different physical parameter. When used with care, multiple regression models can simultaneously describe the physical principles acting on a data set and provide a powerful tool to predict the impacts of changes in the system described by the data.

This article was originally published on Towards Data Science.

Read the original:

What is Multiple Regression? - Built In

Communication, engagement and creativity are key to fostering the next generation of data scientists, said event organiser Aine Lenihan.

Aine Lenihan, also known as Data Damsel, will host a panel of world-leading women working in data science as part of a free virtual event she has organised.

The event, which coincides with International Women in Engineering Day on 23 June, aims to support women and girls entering the male-dominated field of data science.

All genders are welcome to attend the event, which is part of the Women in Data Science (WiDS) Worldwide conference series run by Stanford University in more than 150 locations worldwide.

In her role as WiDSs Irish ambassador, Lenihan said she curated the events speakers carefully with the aim of highlighting to attendees the versatility of a data science career.

Unfortunately, the people shaping the data that shapes the world are a homogeneous bunch AINE LENIHAN

We have something for sports fans and entrepreneurs with the amazing Hlne Guillaume straight from her experience on Dragons Den. We will have the head of data science at Starling Bank, Harriet Rees, as well as TEDx speaker and clinical neuroscientist Rebecca Pope. Well also have the managing director of Accenture Labs, the renowned Medb Corcoran, and Jennifer Cruise, who has just joined EY as director of analytics.

Having mentored and lectured in data and database management systems in Trinity College Dublin, as well as working in the industry for over 20 years, Lenihan is passionate about education and increasing the visibility of women role models for the next generation.

Real-world applications of data science, from combatting cybersecurity to climate change, are shaping todays world and tomorrows. Unfortunately, the people shaping the data that shapes the world are a homogeneous bunch, she said.

Somewhere upwards of 78pc to 85pc of data scientists are male. Bringing women into data science is critical for ensuring accurate and unbiased data is available for todays data-driven businesses.

Lenihans own personal experience as a woman in data science also drives her push for greater gender diversity in her industry. She currently works as a senior analytics architect in IBMs Watson division and previously work in various software and data roles at AIB.

A challenge I have found throughout my career has been missing the company of having other women on my team. Thats a really important factor for me, she said.

Thankfully, there are now lots of active communities, including WiDS, bridging those gaps.

So, what advice would Lenihan give to people hoping to pursue a career in data science and the STEM industry in general?

She predicts the sector will be one of the fastest-growing through to the rest of the decade, and it will be subject to continual evolution of trends in artificial intelligence. AI will also create completely new jobs we havent even dreamed up yet, she said.

My advice is not gender-specific, but a call to those who do not consider STEM for them because they are creative rather than analytical. To those people, I want them to know that creativity is the secret sauce in STEM. Creativity and STEM are no longer chalk and cheese. There is a place for all skillsets and talents in STEM careers. Being creative rather than analytical does not rule out a career in STEM for anyone. In fact, the magic happens when both work together.

Lenihan has herself learned to appreciate the broad expanse of data science and the multitude of backgrounds data scientists can have in her role mentoring second and third-level students on IBMs Pathways to Technology (P-Tech) programme.

As a mentor with P-Tech, students share their career goals and dreams for the future with me, some of which they think are unrelated to AI. But it really gives me the perfect opportunity for me to demonstrate how AI will touch every industry. So you want to be a footballer? Well, AI is being used to create future superstars by boosting performance, minimising injury, predicting recovery time. Want to work as a makeup artist? Well check out the handheld makeup printer, or the NASA-backed skincare micro-mist.

The dynamic Data Damsel who was given the nickname by her former colleagues due to her relatively rare status as a young, female data expert is confident about the industrys future.

There really is so much happening to pique the interest of future data damsels, but communication and engagement is key. There will no doubt come a time I will be more data dame than damsel but hopefully if I continue my advocacy, there will be many generations of data damsels behind me.

Read more:

Women in Data Science event will showcase the versatility of a career in data - Siliconrepublic.com