Search Results

I recently had the opportunity to attend the Open Data Science Conference (ODSC) West - one of the premier conferences on data science and the largest applied international data science conference since 2015. Focus areas include open source programming languages (R, Python, Julia, Scala, etc), the latest Machine Learning (ML) techniques, Predictive Analytics, Deep Learning & Neural Networks, DataOps (machine learning pipelines for production applications), Natural Language Processing (NLP), Data Visualization, Artificial Intelligence business use cases, computer vision, and voice recognition. ODSC touches upon virtually all things data. With so many topics and presenters, you really have to focus on what sessions you need to attend, along with the mindset that any sessions missed will have been recorded. Because of the shift to virtual conferences, most material is readily available after the conference. A great added benefit to this data science conference is that most presenters post their code and material on GitHub. It is very easy to go back and catch material that you may have missed earlier. Slack was a great tool for tracking all the presentations and really allowed me to make the most of the conference. Slack with GitHub is a powerful combination! My intent for this conference was to focus on topics relevant to several professional projects involving Natural Language Processing (NLP), serverless computing (AWS Lambda), and my current school research in Personalized Medicine. My starting point was to look for anything to advance my R/python skills, new natural language processing techniques, an overview of Bayesian statistics, and healthcare Artificial Intelligence (AI) use cases for personalized medicine. I was very happy to find a robust selection within the ODSC West, to include: Health AI: What's Possible Now and What's Hard (Suchi Saria - John Hopkins University) Hands-on Reinforcement Learning with Ray RLib (Paco Nathan - Derwen, Inc) Modern Machine Learning in R (Jared Lander - Lander Analytics/Columbia Business School) Deep Learning for NLP with PyTorch (Ravi Ilango - Stealth Startup) The State of Serverless and Applications to AI (Joe Hellerstein - TRIFACTA) The Bayesians are Coming to Time Series (Aric LaBarr - NC State) Bayesian Statistics Made Simple (Allen Downey - Olin College) I took away several highlights from the conference, to name a few: - Patient data and digital health apps. Connecting patient health apps to electronic health records and physician notes is a current technological challenge. Some researchers are using AI with screenshots from digital health apps to train and develop healthcare-related models. - Reinforcement Learning. I have participated in a lot of discussions as to whether to choose simulation or optimization for various business challenges. Optimization is not very flexible, so I have typically leaned more towards simulation. Reinforcement learning is the combination of these two worlds, and I am very excited to see where this technique can go and how it can be applied. - Serverless Computing. I was not able to watch the original presentation by Joe Hellerstein (TRIFECTA) on the State of Serverless and Applications to AI, but after downloading the material, I saw it was an excellent overview of AWS Lambda, functional computing, its strengths, limitations, and challengers. This was an excellent presentation on this technology and where it is going. I highly recommend this conference to both aspiring and seasoned data science professionals. It is good for both sharpening and learning skills, and for tracking the latest and greatest tools in this space. It was very well done in the virtual format and enabled me to see more material than I could have in person. That being said, I definitely look forward to when we can all connect live at the next conference. The list of all the presentations and keynotes are here: https://odsc.com/california/west-2020-prereqs-2/ In most cases, you can check out their GitHub repositories for presentations. Enjoy! Jerome Dixon Is a Senior Operations Research Analyst at CANA Advisors and can be reached through his LinkedIn profile and via his email at jdixon@canallc.com

In this episode of the CANA Connection Podcast our Host Rob Cranston speaks with long time friend and fellow analyst Rachel Stuve the Director of Data Science from Anthem Health Systems about: Smart Cities, A.I, and Data Science Analytics and how they could affect the future of the Healthcare industry. Along the way they may even talk about their favorite electric powered smart cars, so buckle up and enjoy the show. Link to the podcast episode: CANA Connection Podcast - Smart Cities, Analytics, And The Future of Healthcare. A written Transcript of this Episode can be found below. If you would like to learn more about Rachel Stuve, or follow her lively dissuasions on other podcasts you can find her contact information on Linked in at www.linkedin.com/in/rachel-stuve/ To find out more about CANA Advisors or to talk with our analytics professional head on over to our website at canaadvisors.com. While you are there consider signing up for our CANA Connection Newsletter or joining out community forums. And as always remember; Analyze. Assess. and Execute. Our Host: Rob Cranston (rcranston@canallc.com) Guests: Rachel Stuve (rstuve@gmail.com) Click below for a PDF written transcript of this podcast. Rob Cranston 0:01 Welcome to the CANA Connection Podcast. I am here in Alpharetta, Georgia with a good friend and colleague of mine, Rachel (Stuve), today, I'm so excited about this conversation. It's a long time coming, we're gonna have a chance to talk through and discuss all sorts of cool topics in analytics, innovation, human side of analytics and scenarios that are related to operations. And I think we'll probably get into discussion about our favorite cars. Which, maybe, yeah, which, which is the Tesla versus all other electric car discussion. This is again, a great a great honor to have you here, Rachel, and you've been part of a lot of podcasts. And in fact, you (Rachel) were voted 2019 is top voice data science and analytics on LinkedIn, top voices pretty cool. So let's get at it and give us a background discussion on your career in analytics. Rachel Stuve 0:50 Okay, My background, I didn't practice my elevator pitch. But I'll start so, I did go to school for analytics. It wasn't called analytics at the time, it was called advanced excel and information systems. And then I worked my career always been in data analytics for various capacities, law enforcement, so I actually did do it in jail and all the police cars on at the time it was in Michigan, so boats, police boats, around the Great Lakes. I've been in automotive, manufacturing, healthcare, and that's spanned from startup all the way to a fortune 50 company doing different analytics started out as a programmer, but now on his strategy, and digital transformation. Rob Cranston 1:38 Yeah, so you're now been there for eight months or so? But Director of Data Science? Rachel Stuve 1:42 Yes. Rob Cranston 1:44 Tell us about. Rachel Stuve 1:45 So Anthem, big, huge health care insurance providers, almost everyone's familiar with the shield. They operate in 14 different states, very, very large organizations, and what they're really cutting to operationalize back end. Which people don't realize how complicated a healthcare claim. And there's hundreds and hundreds of data points on any healthcare. And so for even for human that can be really overwhelming to adjudicate a claim when you're looking at that many data points. So what we're really looking at is how we can leverage machine learning to predict claim claim action, how can we route through the system to find efficiencies, which not only save the organization money, but really save members from abrasion. So if you think of times that you've gone to a doctor, and you've had a problem with your claim question or something has to get redone, that can be really arduous process. So we're looking to really alleviate that. Rob Cranston 2:48 So let's talk to these claims. I mean, with COVID. Now, and unfortunately, another uptick in full swing, bear this fall going into the winter is their application of different types of AI that's applied to just I guess, the uptake of different types of claims and claims and you think that could be applied to? Rachel Stuve 3:11 Yeah, absolutely. That's a good question. So I'm also really active in the angel investment community. Yeah. And so really, what COVID has done has drastically increased the peace of digital adoption. So things like telehealth where you would have a video chat with your doctor. That's been around for many years, but people were very slow to adopt that. And COVID has just kind of propelled us a year ahead of digital adoption. So that is a trend that is very likely to stay. I see that a lot too in the startup community. There's a lot of startups that are looking at not just telehealth, but we're so if you have a watch can not only give you feedback, but also send that to your healthcare provider. And so there's been a lot of startups that have started or even that maybe we're not getting much attention that are now getting attention, because when we've all been forced to adopt this quickly, because of COVID. Be sure when the pandemic is over, more people will physically go to a doctor, but there's going to be a good population that are still Rob Cranston 4:22 Going to use all of its advantages. Well, it's interesting, we have telehealth and the patient care needs right now within the teller kind of umbrella of services and offerings. I mean, from a from a standpoint of patient care, do you see that increasing, probably becoming more simplistic right, with access to public say, is that something that you guys at anthem, do you guys have a particular set of software that's services, Rachel Stuve 4:50 So I can't I can't speak to them proprietary methods. But I am a big big proponent, and I've talked with other people about Healthcare isn't just what happens when you go and see your doctor, once a year or twice a year, go healthcare is really every day. And so that's where a lot of organizations in the healthcare space are looking at. How do we deliver health care? How do we make it into the wearables is very large segment, looking at wearable technology, another really huge advancement, where companies are looking specifically with machine learning, natural language processing, and NLP, where you have your phone, and you just talking to say, Hey, where's the closest COVID testing site, and your phone app is going to recognize that and give you a list based on your geolocation, where the closest COVID testing, and it's really that I think that kind of making it really intuitive is where a lot of healthcare organizations are going, and where they're really leveraging machine learning. Where you can even look up, hey, here's the symptoms that I have, what might that be? And, of course, you still need to see a healthcare provider, but it's going to look at that and guide you to that might be best for that type of condition. Rob Cranston 6:19 Right. Yeah, that's great. Well said, I think it's interesting is that continues to evolve as a technology that, especially from kind of a geolocation of those other phones that have an app, it's uploaded under some states that are either symptomatic or asymptomatic, or they're willing to share that information. Yeah, absolutely. Rachel Stuve 6:40 Just COVID, there's a lot. We've talked about this. There's a lot that can be done. Especially with when you have wearable technology, your phone is always with you checking in and saying, Hey, how are you? I read a study recently, where the participants in the study had a phone app. And it just random points in the day they message through the app, and they had to go in and respond to what were they feeling? What were they doing in that exact time. And then what the researchers did was they can analyze that data to look at trends in depression, aggression, you know, loneliness. So there's a lot that can be done outside, of course, where we can really leverage analytics and machine learning, and behavioral in cardiac care. You know, I know wearables that can predict your vitals and say, Oh, my gosh, Here, have a heart attack. Oh, and rushed to the hospital. So there's COVID is talking about today. But there's a lot of a lot of different areas where machine learning can really advance healthcare. You know, Rob Cranston 7:55 It's so this is critical just segues into question. And actually, one of the things that really, when we first met, there was, and this is advertised in a lot of podcasts. And I love this statement. And that is Rachel believes that data empowers humans, is what gives us the ability to solve problems and change the world with data, we can close gaps and move boundaries to become more interconnected with each other. I love that. Oh, that is so cool. I mean, that gives that analytics data over to many countries. Right, that application experiment. A little bit more about this empowering, Rachel Stuve 8:31 Empowering data. Yeah. So one of the the challenges that I have in my career, and this has been across every role that I've had, is, there's oftentimes this fundamental belief that machine learning or analytics or AI is going to replace. And so you come into an organization and we want to do an analytics project or machine learning. And you immediately people put up walls and say, I don't want to be automated. Um, I, I can't tell you how many conversations I've had about machine learning that people reference the movie minority, because they think it's gonna know what I'm thinking before. It's gonna knock doors. Right? We see, we see that in hockey. There's so many movies like Minority Report by robot and all of that, where machines take over. But that's really, the thing that I tell people is, it's data analytics, and machine learning is never going to replace humans, it's actually going to free us up to do what we're better at doing, which is really complicated. And so by leveraging machine learning for some of the more mundane tasks like data, data, cleansing, data, pulling, looking at maybe more mundane patterns, that frees humans up to actually use their cognitive power to make decisions and look at a lot of complex different data points, eventually, with machine learning get there? You know, as a technologist, I want to say yes. But as a human, no, I really don't think that there will ever be a true place. So that's where I say, it really does empower us because it takes that load of our processing power so that we can process more complicated thoughts. Rob Cranston 10:26 So that gets into that. And I can add, in our analytics Operations Group, we have Power BI team analytics. Well, that's the human side of that is the power of having data science ops research, subject matter experts and developers all in one bucket that can deliver this fusion of information. Absolutely. Right. And, you know, so but what gets can be tricky, is the translation of all of that into a client that has to understand English math. How do you really explain an analytics project? The back end of that to make sure that return on investment? So what are some of the techniques you use to make sure that your methods and community and community right, yeah, Rachel Stuve 11:07 So I would say, if you sat and thought about the time that you spend doing is a basic data analysis? So how much time do you spend in Excel? How much time do you spend writing emails, answering questions, or sifting through financial statements? If I took all of that off your plate, and we could automate that and cleanse that you did not have to spend that time in Excel? What would you be able to do? And the answer is, I think when I asked that to people, they sit and think, well, Oh, my gosh, I really do spend a lot of time in Excel, or answering just basic emails, I had an executive that actually measured the time and it was hours every day. And he sat and said to me, you know, if I got those hours back, these are the projects that I could do that work for them. And that's what we use, then as the guidance for where our analytics programs targeted, was, if I don't have to sift through all of my inventory, and I just know that I have does all that time now that I can spend really deciding how to better use my than just, for example, so explaining to people in terms of how it's going to impact them. And what the actual deliverable is, is, is where I'm really stressed. I think, technical people have a tendency at times to also be over complicated, and they use really big words. And I've done that we all want it, because it makes us feel smarter. And you know, we can charge more for that. Um, but you really have to sit and look at it through someone else's eyes of how are they actually using it? And what is it going to see? You, for example, when I was working in law enforcement, one of the biggest things that we did that was I got the most feedback about the whole project was that we changed the visual on the laptop screen to accommodate police officers who. So when you think of a police officer that's outside, Sunny, it might be rainy, there could be chaos, they don't have time to sit and click around and read small print, or if it's funny, that changes how they can view. And it's just changed something like changing the way that your analytics and your reporting was displayed on the screen. Make that project adopted my officers, because they actually it was it was it was absolutely operational. Absolutely. Right. Yeah. Rob Cranston 13:43 That's, that's a great example. Um, so last time I was here, we had a chance, or you invited me, which is fantastic to be part of. It was a smart cities discussion. It was just yeah, that for you. We're Atlanta, integrating different types of smart city concepts. And so my question to you, since then, analytics apply to smart cities. Rachel Stuve 14:03 There are so many ways that analytics can apply to smart cities. So one of the biggest impacts is transportation. And so kind of somewhat, the biggest use case is traffic congestion. And looking at analytics of where were my traffic patterns, where are what times of the day. So you can mitigate congestion, but that also really leads to public safety. So if I can reduce congestion at the same time, I can look at safety. So where are my accident? What type of accidents Should I put around about an inch? I could have traffic lighting, you know, Did someone die? Was that a fatal accident, was it offender. You can look at all of those and start looking at infrastructure. From a citizen perspective. Think of You could tell, we have GPS that reroute. But think about if you could tell how fast the cars were coming if you're riding on a bike, and if you already had a warning that a car was coming, so you didn't cross the street and you stayed into the bike lane. So there's a lot that has to do with safety. There's a lot that has to do with that congestion, but even just moving. So a lot of businesses really can be advantaged by smart cities from a perspective of targeted marketing. So if I know that my customers are closed, I can send them a message to say, hey, you're really close to the Starbucks, guess what it's happening our coffees. So there's a lot that you can do, from commerce perspective, from safety from congestion, a lot with even just think of penances. And moving those through our cities more effectively. Rob Cranston 16:04 Do you think that I mean, those are city planners and you know, those the strategists within cities are, are really what they're building the resilience of strategy for the city that applies to what you just brilliant concepts of being smart and efficient, especially in transportation. But I mean, are they really bought into they have? And is the community in the pockets of kind of broken? Question of like, are there? So part of smart cities worries others? Like, oh, Rachel Stuve 16:34 Yeah, so that's kind of what we talked about is, city planners are in a really difficult position. Because there are a lot of advantages to smart cities, from safety, to transportation, to be more efficient, but typically, residents aren't going to sit and say, Yes, I would like my taxes increased, so that I could feel safer when I ride my bike. It's kind of when you look at even the automotive industry kind of went through the same thing. where, you know, there were extra cost to the vehicle for seatbelts and airbags and anti lock brakes, and all these things that are safety. But when they were first rolled out, people didn't want to pay for that. Because what do I need an air? Right, I'd rather have some really fancy wheels. But over time, the automakers had that challenge of adding that safety and that cost in and trying to still balance what consumers would pay for. So from and with the automotive industry, a lot of these really kind of driven by regulation, adoption. So that's where I think city planners have been tough. But that's where I think us in the infrastructure and data industry, that's our job to be kind of evangelists, to people to say, here's the benefits. First, why we might want to one of the really great case for smart cities is public safety, right? Not only can I get my officers more quickly to an incident, but maybe I can actually preemptively place officers where I think there might be Rob Cranston 18:15 And predictive, a little bit of predictive. Rachel Stuve 18:17 Bit of predictive. And that's kind of another story, because there's ethics around that kind of conversation. But do you find what you're looking for? But I think that's where there is a lot that could be an advantage to a smart city from residents. Absolutely. From commerce, absolutely. From logistics. But when we think about safety, efficiency, there's a lot of good cases. We'll get there. It just takes more conversations. Rob Cranston 18:47 Yeah. And that's what we're gonna do next time. You're talking about the automotive industry. So would you tell everybody what the name of your car is? Rachel Stuve 18:55 I'm a huge Tesla fan. We have we have sparred quite a bit. I did, did shed a tear when you bought that Audi. Rob Cranston 19:01 Yeah. And for our audience out there. Rachel does have a Tesla, has a redTesla hat on. But it's a beautiful, beautiful car. And it just, I think, in general, getting back to the discussion of, you know, efficiencies and that smart technology. Most of these companies, although it has a bet ter battery, I mean, then a Tesla... (laughter) Rachel Stuve 19:23 For the audience that can't see I'm very upset by that comment. (laughter) Rob Cranston 19:27 Well, I love it. You are an incredible mind. Huge leader within the analytics community and just the continued growth of where analytics is gonna go. I appreciate you being on this podcast I look forward to more. Rachel Stuve 19:41 Likewise. Yeah, I know CANA is doing some really phenomenal things with not just analytics that can be leveraged today, but I really am impressed with what you're doing to set people up for the future. Like my Thank you for having me for this conversation. All be it a Tesla is better than Audi by a... Rob Cranston 20:05 That'll be what we'll go into. That'll be our different Rachel Stuve 20:08 Different conversation. Yeah, Rob Cranston 20:09 Yeah, we just have those two (cars) and a podcast. Right. Yeah. The facts. Rachel Stuve 20:13 Yeah. I don't know if you wanted to be an analyst by nature... Rob Cranston 20:20 800, I mean 600 lbs heaver. I got a lot more space. Rachel Stuve 20:23 A lot of data. Rob Cranston 20:25 That'd be great. Awesome. For your audience. Is there anything to share? Rachel Stuve 20:29 Yeah, for my audience, I recommend people reach out on LinkedIn is the best platform where I've got some other podcasts published and messages. That's really probably the best way to get a hold of me for good conversation, questions. Always, that's the best way that I can. Rob Cranston 20:49 It's so responsive. Holy cow. I see it firsthand.

Today is a celebration of our military veterans. It is both a solemn and joyous day - a continuation of the commemoration of Armistice Day, marking the end of hostilities in the first World War. Although that “war to end all wars” was not, in fact, the end, it was the beginning of a continuing respect and gratitude for the American military veteran. President Dwight D. Eisenhower officially proclaimed November 11 as Veterans Day in 1954 with the following words: ….On that day let us solemnly remember the sacrifices of all those who fought so valiantly, on the seas, in the air, and on foreign shores, to preserve our heritage of freedom, and let us reconsecrate ourselves to the task of promoting and enduring peace so that their efforts shall not have been in vain…. CANA Advisors honors all veterans today, 11 November 2020, and we try to dedicate ourselves to that obligation throughout the year. To that end, we are so very honored to be recognized by the U.S. Department of Labor as a #HIREVets Medallion Award recipient, for a commitment to attract and retain veterans to the workforce. We take this recognition seriously. CANA celebrates this Veterans’ Day knowing the entire nation has immense pride in, and admiration for, our American military veterans. #veteransday #hirevets #teamcana For more information on the U.S. Department of Labor's HIREVets program go to https://www.hirevets.gov/ To learn more about how CANA Advisors is involved with our veterans go to https://www.canaadvisors.com/

The Problem The R programming language is a popular open source environment for statistical computing. CANA runs multiple R algorithms on a server using the Plumber framework, which converts existing R code into web APIs by adding a couple of special comments. A key issue with R is that it is single threaded, meaning that only a single block of code, inside an application, can execute at any given time. In other words, while that single block is executing, all other code is blocked. This is not typically an issue unless there is a block of code that takes a long time, such as a complex algorithm or a call over a network. CANA’s application performs both, and supports multiple users concurrently, so being single threaded does not scale for us. One solution I pursued over the last year was to use the AWS Lambda service which allows standalone functions to execute in the cloud. The challenge with Lambda is it does not natively run R code and is better suited for Java, Python, and other languages. While it is technically feasible to run R inside of Lambda, we had not yet implemented it because it is non-trivial, requiring that R run inside a Python container with supporting frameworks and libraries. The Solution A coworker came up with an elegant solution to this programming problem. The solution is to run a Node.js server that simply makes calls to shell scripts (R scripts) running on the host operating system. Each shell script runs in its own process, so one won’t block another. And, even though Node.js is, itself, single threaded, it uses non-blocking input/output calls (“callbacks”), allowing it to support tens of thousands of concurrent connections without incurring the cost of thread context switching. Simple, elegant, and trivial to implement. Joseph Moreno is a Director of Development at CANA Advisors. You can follow him on his website at joemoreno.com or contact via email jmoreno@canallc.com

As a Senior Operations Research Analyst at CANA Advisors, I’ve had the opportunity to apply both my military logistics experience and healthcare analytics expertise to a variety of challenging problems. I recently shared a learning experience with peers at our monthly CANA Analytics Roundtable (CAR). The CANA “CAR” is a monthly gathering open to all hands of the company, which highlights employee-chosen topics related to analytics techniques and technologies. Each monthly CAR typically hosts four to five short presentations. During our most recent roundtable, I talked about my strategy and initial execution of a cancer trial data science challenge hosted by Oak Ridge National Laboratories. The challenge set forth was to analyze cancer patient information and data sets to appropriately determine an individual’s assignment to a select clinical trial. I identified 25 input features within the patients’ information database that would eventually feed into the target variable – “Selected for Study, Yes or No.” I noted a specific need for improvement based on doctors’ feedback: trial names did not always match to the relevant disease site, thereby missing a critical linkage point between patients and potential trial participation. I used PyTextRank as one means to address the data challenge. Although the “bag of words” is a common model in text mining, it focuses mostly on simple word identification and count. In this instance, I felt Python’s Pytextrank library was the right tool, given the types of trials and abstracts in the challenge, to classify titles to the correct cancer sites. Pytextrank can be used not only for identification and counting but also to select keywords, assign importance to the word, and build summary sentences from text. I used Pytextrank to review the summaries and to establish an Eigenvalue centrality metric that ranked node importance based on not only the number of connections but also the quality of the connected nodes, essentially creating a network strength metric. Another critical tool was the R deep learning API, Keras. Although it appears a difficult language to work in, it seemed most of the heavy lifting work was done in data preprocessing to put the data in matrix vector format. Keras was critical in addressing my intent to define and train the model to appropriately match a large body of cancer trials to specific cancer anatomical sites, e.g., brain, breast, prostate, etc., thereby enabling efficacious patient match-up to a potentially useful trial. In order to put these different elements together, I used Reticulate to embed Python Pytextrank in R. This approach was fairly effective, and I was able to demonstrate initial iterations of my model. As I continued through validation and analysis of my approach, I realized the classification model did not produce what I considered significant results. I need to further feature engineer the text corpus dataset and improve the model's input features. This iterative process will help determine features that best represent, classify, and connect the data flowing into the model to provide optimal results. My next steps are to experiment and test out the methods used here in https://cloud4scieng.org/2020/08/28/deep-learning-on-graphs-a-tutorial/. This deep learning approach may reveal more about the underlying structure of the cancer study data; define the nodes and edges that detail its connections and features; identify or predict links and communities; and enable classification between classes. I intend to, quite literally, connect the dots of the data to solve this cancer clinical trial challenge. Jerome Dixon is a Senior Operations Research Analyst at CANA Advisors. jdixon@canallc.com

In this special Team CANA spotlight episode, our host Rob Cranston will be talking with a handful of active military and veteran TEAM CANA members, about some of the special military workforce transitioning programs that are available, how to get involved in those programs, what it's like making the transition to the civilian workforce, and how the team at CANA Advisors and the CANA Foundation are involved. A big thanks goes out to Hannah Wallace, Walt DeGrange, Cornelious Young, and Kenny McRostie for being part of today's Team CANA veteran spotlight interview. Link to the podcast episode: CANA Connection Podcast - The Transitioning Veteran Workforce If you would like more information about the programs mentioned in the show or how you can get involved, hit up our website and specifically our CANA Foundation page under the about section at canaadvisors.com. While you are there why not consider signing up for our newsletter, checking out some of the blogs, or joining our community forum. The CANA Connection podcast is available on Spotify, and on many of the other popular podcast Streaming services. You can also find us on all of the major social media platforms at CANA Advisors. As always remember, Analyze, Assess, and Execute. Our Host: Rob Cranston (rcranston@canallc.com) Guests: Hannah Wallace (hwallace@canallc.com), Walt DeGrange (wdegrange@canallc.com), Cornelious Young (cyoung@canallc.com), and Kenny McRostie (kmcrostie@canallc.com) Links DoD SkillBridge Program https://dodskillbridge.usalearning.gov/ Hiring Our Heroes https://www.hiringourheroes.org/ The CANA Foundation https://www.canallc.com/giving-back CANA Advisors website https://www.canaadvisors.com

What is Mural and how do you use it? Great question! Mural is a virtual collaboration whiteboard you can use anytime and anywhere. It is the perfect platform for CANA’s virtual team members who are located throughout the United States and Canada. With such a widespread group, finding communication and collaboration solutions is key to our success. Mural is able to import website links, documents, photos, and graphics. Mural is one of CANA’s project management “go to” tools for collaborating on-line with clients and the CANA team. Our project managers can create templates to design, plan, and execute tasks for a project based on multiple contributors’ input, synthesize the collected data, and reach a decision faster than using other traditional techniques. Mural has proven to be outstanding for virtual collaboration and getting every person on the same page. A recent example of a CANA project use case Originally, CANA selected this platform to plan and execute the Future Unmanned Logistics Systems (ULS) Energy Logistics Enabling Distributed Operations (FUELED Ops) Virtual Demonstration Day and Design Think Workshop. We needed an on-line platform to collaboratively and simultaneously capture multiple users’ inputs into event design, planning, and coordination. We used Mural to visually map out and plan the event. This gave the planning team a visual representation of the event and kept all of the key information in one place. This dashboard had 16 different planning boards, and one of these included a storyboard used to plan out the narrative for the virtual event. As a virtual team, it was very important to put together the client’s vision in a way they could readily understand. It increased client confidence that the event would be successful, and their story clearly and accurately conveyed, to the event’s audience. This is the storyboard mural board the FUELED Ops team used for the event’s dry run. This event showcased the market research and modeling and simulation and provided an unmanned system demonstration for the FUELED Ops Program’s Year 1 efforts. In addition, during the event, CANA held brainstorming sessions to bring concepts into reality. The Mural board display below was an activity we conducted to come up with a common operational view. The board on the left shows the ideas for the “game pieces” and what we thought the operational concept should look like. The final result, on the right, is a 3D island model and includes game pieces created by CANA’s Graphics Artist. The team was able to use the game pieces and copy and paste them into the island visual to create several operating concepts. This provided an exciting and interactive experience to the FUELED Ops participants. CANA also hosted the FUELED Ops Design Thinking Workshop after the virtual demonstration. The information from the event will feed into the FUELED Ops Program Year 2 efforts. CANA was excited to host this event, having understood the challenges of how to achieve the best results in a virtual environment. Below are the morning and afternoon session activity boards used to diverge and converge on the future of unmanned logistics systems in the military. A few examples of CANA operations use cases A great feature of Mural is that it is completely customizable, and a team can create the board in a way that will work for them. Two great CANA examples are our lessons learned activities and the on-boarding of new employees. We conduct lessons learned activities on Mural to have meaningful conversations in a judgment-free, neutral zone. As a team, we are constantly looking for ways to improve our communication, and Mural allows us to create a space for everyone to add their thoughts and have open discussions. The tool also has a voting feature that allows us to vote on important issues and prioritize them. Another example is the CANA new employee on-boarding Mural board. We created this board to conduct an easy first-day orientation with documents, presentations, files, and more to help our newest employees understand what to expect for the first day, week, and month. Mural makes it easy to customize these boards for each employee - we can publish a template and customize it specifically to the individual. CANA’s newest Hiring Our Heroes Corporate Fellow, Neil Young, recently used our on-boarding Mural board, and he commented, “[c]ompleting my onboarding using Mural has been extremely helpful, especially with having diagrams, training, contact information, documents, and the ability to collaborate on tasks in real-time.” Mural works well for CANA, could it work well for you? Mural has helped our virtual company collaborate with, and understand one another, in a way much like being in person. It's a very diverse tool kit full of unique features that fit our specific needs, and it works well for CANA. We continue to build new use cases every week using this tool. Does your company use Mural? If so, we would love to hear how you are using the platform! #agilebusiness #mural #canaadvisors #teambuilding #collaboration #communication Hannah, a key Project Manager at CANA Advisors, played an instrumental role in designing, planning, and coordinating the FUELED Ops event and was pivotal to its success. She brings curiosity and commitment to CANA's project management practice - as demonstrated by her research and assessment of MURAL as a force multiplier to our client engagements - and she is a leader within our team often engaging and implementing new technologies and efficiencies. If you'd like to learn more about Hannah Wallace or CANA's project management practice, please reach out to her at hwallace@canallc.com.

Be a problem solver, not a problem maker. - Unknown Todd will serve as a Principal Software Engineer, Development Lead and Information Assurance (IA) Lead for CANA.  He has a diverse background in the software, systems development, information technology and cybersecurity fields.  He has 23 years of experience in the defense, intel, commercial and financial sectors including system design and implementation, project management, bid and proposal, information security, infrastructure planning, technical leadership and strategy roles. Todd holds a bachelor's degree in Electrical Engineering from the Pennsylvania State University as well as a Master's in Engineering Management from Robert Morris University. You can contact Todd at tallison@canallc.com Just because something doesn’t do what you planned it to do doesn’t mean it’s useless. - Thomas Edison #teamCANA #CANAAdvisors #CANAConnect CANA Advisors is a veteran-owned, woman-owned, equal opportunity company based out of Gainesville, Virginia in the United States of America.

“Failure is not an option…. Either I am winning or I am learning, never failing!" - Nelson Mandela We are excited to bring aboard our newest HoH Fellow, Cornelius “Neil" Young. Neil will serve as a Project Manager for CANA and is a highly regarded U.S. Air Force leader with more than 23 years of proven success in healthcare operations, strategic leadership, management, and planning. Neil has experience and expertise spanning all facets of healthcare delivery and program management, to include training, fiscal year budgeting, and the deployment of healthcare assets to support contingency operations. Neil holds a bachelor’s degree in business and an MBA with a specialization in healthcare administration from St. Leo University. You can contact Neil at cyoung@canallc.com. Pictured From Left to Right: Rob Cranston (CANA President, COO), Jason Fincher (CANA Principal Logistics Analyst), and Neil Young (Hiring our Heroes Veteran Fellow). #teamCANA #CANAAdvisors #CANAConnect CANA Advisors is a veteran-owned, woman-owned, equal opportunity company based out of Gainesville, Virginia in the United States of America.

In this post, I take a look at bicycle traffic patterns in Seattle, Washington, in light of the ongoing COVID-19 pandemic. Also, given very recent events and the timing of this article’s publication, I take the opportunity to include updated data for June, which presents bicycle traffic patterns affected by the recent social protests in addition to the pandemic. The data supporting this analysis was sourced from the Seattle Department of Transportation Bike Counters webpage, which registers daily bicycle traffic at 12 locations throughout the city. The original analysis was inspired by an R for Data Science Tidy Tuesday community event on GitHub; I highly recommend checking out the Tidy Tuesday repository for interesting, quick, data challenges. Seattle Bike Traffic Overview First, let’s take a look at an overview of tracked Seattle bike traffic over time. A few high-volume crossings (e.g., BGT North of NE 70th and the Elliot Bay Trail) stand out as well as some gaps in the bike counter data. We lost access to the X39th Ave NE Greenway counter in mid-2018 with no visible return. There is also clear seasonality to this data, as might be expected given Seattle’s rainy winters. Using NOAA weather files for the area, we can take a quick look at the relationship between the number of bicyclists on the road and area precipitation. We consider a “rainy” day to be one with at least 0.5 inches of precipitation. We can see a relationship between precipitation and bike count as well as a relationship between temperature and bike count. Bicyclist counts are noticeably higher on sunny, dry days than cold, wet ones. In Seattle, rainy day bicyclists are likely regular commuters. Next, let’s look into the hourly bike traffic patterns at different crossings to get a sense of their usage. Note: This graphic was inspired by a live screencast by David Robinson. Check out his YouTube channel for more Tidy Tuesday analyses. Many bike crossings show clear commuter patterns on the weekdays, with ridership hitting its highest counts around 9 am and again at 5 pm on Monday through Friday. For most crossing locations, weekend traffic peaks around noon. COVID-19 and Bike Traffic Seattle was the first major U.S. city to be affected by COVID-19 starting in February, 2020. How have the imposed social distancing practices impacted overall transportation trends over the past months? First, let’s look at the overall growth in COVID -19 cases compared to daily bike traffic. With pandemic isolation practices in place, there appears to be a cyclist traffic peak in April and early May. In late May and June, we start to see a decrease in these cyclists as businesses begin to reopen. How, then, has the pandemic impacted commuter patterns? As the pandemic worsens, we see a loss of commuting structure to the bicyclists’ movements over the course of the day. By April, the transportation pattern differs significantly from previous years, with counts peaking in the afternoon. Finally, let’s look at these cycling patterns broken out by counter location for the month of April, in comparison with average April ridership patterns from 2014 through 2019. These quick snapshots of Seattle’s bike lane traffic show a changing picture of transportation in the midst of a pandemic. As social distancing practices make public transit or ride-hailing inadvisable modes of transportation, many people turn to biking for both necessary transportation and leisure activity. Local Social Protest Activities and Bike Traffic Seattle was also the site of some of the nation's largest organized social protests in response to the May 25th death of George Floyd. Notably, the Capitol Hill Organized Protest (CHOP) created and occupied an autonomous zone for several weeks in June within Seattle’s Capitol Hill neighborhood. One of the bike counters we studied in this analysis, the Broadway Cycle Track N of E Union Street, is located on the edge of this zone, providing a unique look at movement during this time. As the community responds after the death of George Floyd on May 25, 2020, we see a sharp drop in cyclist movement near this counter. When the autonomous zone is established on June 8, 2020, movement falls close to zero. Unfortunately, we can’t see the full picture with regards to pedestrian data at this location and time. However, the bike data shows a clear change in movement patterns and gives an indication of the major impacts of this event. In Summary Events of the past several months have been unprecedented in scale and scope, with tangible effects on the micro- and macro- level. Changes at each level may provide valuable insight into the impact of events throughout the world. How have COVID-19 and social protests impacted transportation patterns in your neighborhood? Have you seen follow on effects outside your community? Share your comments below or in the CANA Forum. Lucia Darrow is an Operations Research Analyst at CANA Advisors. If you would like to learn more about Lucia and CANA Advisors’ upcoming courses in analytics or the CANA Foundation, please contact her at ldarrow@canallc.com

“What’s next?” was the ultimate question I heard last year when I graduated from college. It is a daunting one. Now, some of you may have jobs lined up, plans to travel the world, or maybe you have no clue what is next. That is perfectly okay. We all want to rush into independence, rely upon ourselves, and make money; but I have learned this - now is the time to really find out about yourself and what you need in life. Some of you may have come to this article because you are looking for what is next as a data analyst. How are you going to make your way in the world of analytics? Is it right for you? There are a lot of unknowns, but one of our CANA team members and Director of Analytics Capabilities, Walt DeGrange, is here to help. He broke down the life of typical day as an analyst to give you insights into what this job entails: Research - 10% Keeping up with the art of the possible is a required daily chore. New methods and technologies are being introduced daily. Assuming a software math solution implemented six months ago is still state-of-the-art is risking irrelevance. Coding - 30% This is the basic skill required for all analytics professionals. As important as the carpenter's tools, coding in various languages such as R, Python, C++, and SAS allows the analytics professional to manipulate and gain insight from data sets. Communication - 25% Communicating with collaborators, clients, project leads, and technical experts is critical to ensure that deliverables are on time and fulfill the requirement. Marketing - 15% Everyone needs to sell. Even the coder that never presents to a client must convince their project lead that their methodology works. This is a very important skill for analytics professionals since many models use math that is not easily understood or explained. These "black box" solutions require a higher level of convincing. Project Management - 10% Keeping analytics projects on track is not like managing a construction project. There are many analysis areas that require familiarization with the data before building a model. Many aspects of model building are more of an art form than a science and thus the time to complete may have a large variation from project to project. One must consider this in the planning and execution of these projects. Breaks - 10% Everyone needs a break, and this is especially true if your job keeps you in front of a computer screen. Walking, running, and cycling gives you time and space to think about challenges. Sometimes your subconscious mind needs this distraction to develop solutions. Plus, the physical exercise is just good for you. Walt sums it up this way: “Of course, this is just a sample day. Why I love analytics is that I can apply the techniques across many industries and solve a multitude of challenges. This results in schedule variation every day.” I know each of you will enjoy the challenges and excitement of your chosen field, whether in data analytics or elsewhere. You have earned your college degree, and we here at CANA applaud you! Congratulations!

To our friends, colleagues, and partners - Words feel inadequate to describe the COVID-19 pandemic moment that has encircled our world, our communities, and our lives. On behalf of the CANA Advisors team, we wish you and your families health, safety, and optimism during this uncertain time. We are a resilient human race, and we shall prevail.   Yet, even during a time of national and world-wide crisis, silver linings trickle through. Working from home, now fondly hash tagged #WFH, has sprouted – nigh, it has vaulted – into our every day lexicon. For many, this new normal of working remotely has challenged them in good and not-so-good ways. Here, at CANA Advisors, we have been working remotely, or as we like to dub “virtually,” since our inception over ten years ago. Headquartered out of the Northern Virginia area, our team works virtually throughout the United States from San Diego, California to Hollis, Maine down to Albany, Georgia and back to Denver, Colorado!   This quarterly issue has proven timely as we share CANA perspectives and professional development activities demonstrating the strengths and opportunities found in working “virtually.” We also highlight one of our incredible CANA team members – Ms. Kaitlyn Wark – a professional Data Scientist with young children, who is owning this #WFH moment.   Finally, we want to take a moment to thank our country’s first line responders – doctors, nurses, aides, hospital administrators, police men & women, fire fighters, paramedics and every person stepping up to assist when needed – as well as our logistics backbone – the truck drivers, the railway conductors, the postal service, our third party logistics carriers, and every person who is stepping up to ensure much needed supplies and resources get to where they need to go. We salute you! ~Team CANA Reflections on rstudio::conf 2020 RStudio February 2020 by Lucia Darrow This January, several team members from CANA Advisors attended the RStudio Conference, both in person in San Francisco and remotely! At CANA, the R language is a part of our daily analytics practice, from quick turn analyses to high impact applications built in R Shiny. The RStudio Conference provides a great opportunity to learn about the latest trends in the language and connect with others in the community. Here are a few takeaways from CANA Senior Operations Research Analyst, Lucia Darrow, who attended the conference in San Francisco: Virtual Attendance  As a fully remote team, CANA Advisors is accustomed to making the most of virtual meetings and seminars to share knowledge and build community. This conference was the first time several teammates attended talks online simultaneously, coordinating to make sure we covered all the topic areas of interest. The conference organizers allowed virtual attendees to ask questions in the main queue and participate with talks as if they were in the room. It was awesome to feel as if the other R users on our team were attending the conference with me, catching the talks I wasn’t able to attend! In light of the postponement or virtual transition of many conferences scheduled for spring and summer 2020 due to COVID-19, creating the capability to attend and engage with virtual events becomes even more important. Trends and Best Practices Two trends that seemed to be of growing interest amongst intermediate and advanced users are custom R packages and taking R into production. With a slew of new tools and packages to make package creation easier, the use of internal or project specific packages has increased drastically. Formalization of shiny applications into production grade apps was a continuing theme this year. New packages like golem show promise in making this process more attainable for R programmers lacking traditional web experience.  At CANA, we are always looking for best practices to improve our R products and streamline the connection to other tools. I enjoyed several talks in the Programming track which elucidated topics like parallel computing and asynchronous processing with R. As R is a single-threaded language, these methods become increasingly important for processes with longer run time. R Community and RLadies In keeping with their commitment to the open source community, RStudio announced their new status as a Public Benefit Corporation. Diversity and inclusivity were promoted throughout the conference, including a RLadies meetup event hosted by RLadies SF! This was a great opportunity to connect with other chapters and share tips and tricks for building local programming communities. Interested in learning more? Many of these talks are available online. Check out https://resources.rstudio.com/rstudio-conf-2020 to experience the conference. To view CANA’s take on last year's 2019 RStudio Conference, check out Lucia’s previous blog post. Lucia is a Senior Operations Research Analyst at CANA Advisors. To find more content on our favorite professional events, continue to visit our CANA Connection. TEAM CANA MEMBER SHOWCASE Kaitlyn Wark | Data Science Analyst CANA team member, Ms. Kaitlyn Wark, is a data science analyst for CANA Advisors. She develops predictive models and data analysis software that empower customers to make data-driven decisions. Kaitlyn has six years of analytics experience in the public and private sector and leverages domain expertise in education, housing, non-profit management, and insurance. She is a board member of Mentor Virginia. “I've learned that people will forget what you said, people will forget what you did, but people will never forget how you made them feel.” -Maya Angelou Kaitlyn specializes in transforming technically complicated analytics products into understandable, visually attractive applications for non-technical stakeholders. Since joining CANA, Kaitlyn has taken on the role of supporting our expanding data science practice. Previous to CANA, Kaitlyn was a Homeless Projects Evaluator at the Department of Behavioral Health and Developmental Services (DBHDS). Kailtyn architected a data exploration and reporting application that allowed DBHDS leadership to quickly assess the overall performance of statewide housing programs and communicate areas of improvement to local housing programs. Equipped with evidence of the program’s impact, DBHDS successfully made the case for $9.2 million in continued funding for PSH programs to the General Assembly. In her role as Program Evaluation Officer, Kaitlyn developed The Community Foundation for a Greater Richmond’s first interactive dashboard, thereby enabling the Board to easily monitor the results of the Foundation’s social investment strategy and shift towards data-driven grant-making. Kaitlyn is passionate about applying software development best practices and design principles to data science software. In her non-profit and state government analyst roles, Kaitlyn championed the use of open-source statistical software to automate routine analysis and reporting tasks. Kailtyn also has experience deploying analytics software in a larger-scale production environment. She designed and deployed predictive models at Markel, a Fortune 500 insurance company, helping underwriters to more effectively and profitably manage their time pursuing new business. Kaitlyn introduced software development principles and successfully advocated for a test-driven development approach to the team’s data science workflow to increase the scalability and technical rigor of the analytics products they shipped. She contributed to a small team’s ambitious effort to develop software that scaled the development and deployment of customer churn models across multiple lines of business. This scaled approach reduced the time required to spin up a new customer retention model by 67%. Kaitlyn holds a Masters in Social Service Administration and a Bachelor of Arts in Sociology from the University of Chicago. One thing we take pride in is our CANA Culture. As Kaitlyn says, "I love the camaraderie and can-do culture at CANA. All of my colleagues are excited by challenges and naturally curious, and that makes work fun." Working from home is a routine for every CANA member, and Kaitlyn's favorite part about WFH is "I love that I get to trade my commute time for breakfast with my kids and extra play time in the back yard before dinner. I also love that working from home puts me in charge of coffee selection and the brewing pace!" If you would like to learn more about Kaitlyn and how she and CANA Advisors can take your Data Science Analytics to the next level, please contact her at kwark@canallc.com! #teamCANA #canaadvisors #kaitlynwark There Is No Perfect Way to Work Virtually, So Stop Looking for One and Start Working! Thoughts On Working Virtually in Today's World by Walt DeGrange There is no, one right way to work virtually. This may not be a surprise to folks that were forced into this mode by the recent Corona Virus situation. Of course, you can do a Google search and get thousands of "Ten things to make working from home easier" or "Increase your productivity working from home." You could literally spend the next ten years surfing these links, but I would argue that you would see marginal improvements at best and waste tons of time and money. So why do I know this, and what should you do? First, I have been working virtually for six years. Over that time, I lived in a two-bedroom apartment, many hotel rooms, and a three-bedroom house with a dedicated office. I have worked virtually on the road from airports, coffee shops, co-working facilities, and baseball stadiums. So I have experience. The second question is, what should you do? My short answer is, "get to work!" Let us first explore the non-virtual workspace, better known as the office. Offices are basically factories for knowledge workers. Spaces are designed for efficiency with team members located close to each other. Conference rooms, bathrooms, and snack areas are centrally located to allow for quick use. Organizations design office space to increase productivity and minimize cost. Variation is the enemy of a well-designed office. Your virtual workspace is designed by you. It can be a space designated as a home office, a lawn chair on your back deck, a couch in the family room, a kitchen table, or a seat at a baseball stadium. The sky is the limit, and the variation is infinite! That is both good and bad news. After all of those years, not having to think about the space where you work now, you have choices. To say this might be a little overwhelming is an understatement. Where do you start? I would recommend just start with what you have. You have a kitchen table, use it. A dedicated room for a home office, use it. You just need to start working. As you work over the next few days and weeks, you will begin to make observations. Perhaps your kitchen chair is uncomfortable after one hour sitting in it. Put any observations you make on a list either on paper or electronically. After one month, review your list. Then develop a list of ideas to mitigate your negatively impactful comments. You can list buying something to make things better, and I also challenge you to list ways to fix things by not spending money. For example, that uncomfortable kitchen chair, obviously you could buy a pad for the chair or perhaps a new, more comfortable chair. You could also get up and walk around the room every 45 minutes and take a break or change locations completely. Every person's experience working in the virtual environment is going to be different. There is no kit that organizations can send all of their team members to work efficiently virtually. Therefore sending all team members a standing desk is not going to create a great work environment for everyone. There is too much variation in team member's virtual spaces (i.e., house, coffee shop, co-working facility, etc.) to create a one size fits all solution. You must discover what works for you, and that takes time. Interested in learning more? We will be talking more about working virtually in our upcoming free webinar series. Information is below.  Walt is our Director of Analytics Capabilities at CANA Advisors. To find more content on working virtually, continue to visit our CANA Connection. #WFH #WorkingVirtually #VirtualCompanies Connect with CANA ADVISORS in 2020 UPCOMING EVENTS CANA Advisors invites you to join us again on April 16th for Part Three of our webinar series on Working Virtually. Part Three is a conversation on parenting and caregiving while working from home. Our team members will share their stories, challenges, and tips in their roles as caregivers and professionals. SIGN UP HERE Sign up for the CANA Connection Newsletter! CANA Advisors 7371 Atlas Walk Way Gainesville, Virginia 20155 Telephone (703) 317-7378 Facsimile (571) 248-2563 Privacy Statement | Subscribe | Unsubscribe The CANA Connection Newsletter Copyright ©2019 CANA LLC. CANA Advisors is a veteran-owned, woman-owned, equal opportunity company based out of Gainesville, Virginia in the United States of America.