 Everybody, hi, we're going live. I'm just going live a little early, just so everybody can get a chance to connect. And I can get a chance to make sure that I'm squarely in the middle of the screen. Hopefully you'll be able to see the chat comments. Actually, I'm gonna just make a chat. Hi, everyone. Can't post to some channels. Oh, he says that. Chat overlay. Oh, yep, there it does. It says hi, everyone. Okay, good. So that means you can see my chat. And if anybody wants to say hi or anything or say where you're from, where you're listening from or who you are, maybe I know you, go ahead and start doing some chat because then I can see if it was working. I was having a little trouble with this restream. You know, interestingly, this software contacted me and asked me if I'd be in a research study. So I'm gonna be in a research study about this software that I'm using called Restream to do the live stream. But anyway, so that way I can, maybe I'll even get to know more about the software, but. Anyway, we'll just wait and see if anybody joins. I'm so jealous of these casino streamers that go to the casinos and they play slot machines. And there's always like 700 people on their channel or 1,000 people on their channel. And I'm always like trying to get a few people on my channel, but you know, everybody's spreading the word if you've got data science questions or you're not feeling good about data science or you're feeling really good about data science, you know, come to the chat and discuss it. You know, especially the problems that it's hard to discuss like at work, you know, because if you discuss them at work then people know you don't know that or something. Like I used to be married a long time ago and my husband used to be like all sketched out over posting on Stack Overflow or places like that. We used to have this thing called use net where you could post to ask questions. I think it was use net. And he'd always be like, I don't wanna post on there because people will figure out who I am and that I don't know how to do something. And I don't know, I don't know. I always, I never really worried about that but then I always worked at the government. Nobody knew how to do anything at the government. It felt like, you know, I mean, actually people at the government knew how to do stuff. Like there's some of the most smartest people. It's just that trying to find the person to answer your question who knows about it. That's the challenge at the government that I used to have was just finding, just finding the right person who knows the right thing. But, you know, that's part of it. That's the informatics part of data science as the communication part, finding the right people. Well, I'm just talking a little before we get started because today, you know, and I just plan on giving an update. You know, I was watching, there's these telehealth people, they give updates, right? And their updates are really short. They're like five minutes or something. And I'm like, well, I have more to say than that. But, so I made a LinkedIn learning article with some of the updates, but I'm gonna do more because I've been, the reason I haven't been live streaming a lot is I've been writing a lot. I've been writing a lot of peer reviewed articles. And also just my usual tutoring and mentoring and stuff. But, when I write a lot of peer reviewed articles, it often means I'm doing analysis. And so then I have some new code and new thing, new insights that I can share with you. But that takes some time to put together a blog post with code or a video with code. So for now, I just thought I'd give you an update and tell you what's going on. Maybe you can tell me what's going on because there's so many different parts of data science in our field that it's hard to tell what's going on. So it's good to have an update. And there's stuff I didn't even put in that article. Like, because I remember I went to two different webinars and they were both good. They were both on different top data topics but they were both good. You're really right about those. Okay, well, it's noon. So I'm gonna officially open the chat. Welcome to live chat, data science update and discussion. Get your chat on. And so before we go too far, I'm Monica Wahee and I've got this chat overlay here in Restream. And supposedly we should be able to see everybody's chat but I've been having some trouble with the software. So, so sorry. So those of you who know me, welcome to our data science chat. Those of you who don't know me, I'm an epidemiologist, that's a data scientist also. I have a blog and I have my, I'm mainly trying to build up my YouTube channel but I'm also a LinkedIn learning author and an author of textbooks. But also like I just have a business. I do consulting, I do data science. I teach data science, I mentor data science. And what does that mean? Well, that means that sometimes I get really busy with my customers who need data science and then I can't do a live stream or I keep up with my blog or all the other problems that other people also have, right? So, but the good side of it is, as I would say before, I officially started the stream is that when I do work with customers, real life customers, I learn like what's going on in the industry and I learn about what people's issues are nowadays. And so I like to share that kind of stuff with you because then, you know, because a lot of people, a lot of my learners are trying to move into a new career or try something new. And also the world around us is just changing so much. So, so I, there's a lot to update you on but I put some stuff into this article that I put on LinkedIn, which I'm gonna share my screen now, kind of look at some of the things in there. And feel free to, let's see here. I don't know which one of these I want to see. Okay, so this should be it. So this is the, and this was in the link in the description. So this is just the article I wrote. And so the first thing I want to say was, you know, I was so busy through February and so far in March that I kind of was missing Black History Month and Women's History Month. But I wasn't really missing it inside myself because I was thinking of the women that are inspiring me that I want to inspire other people that I want to share their science with others. And so there's these three women I want to highlight and the reason I want to highlight, well, there's sort of two things. There's the sort of basic thing of I want you to know about their research and what they're doing, because the topic is really important. But secondly, I also want you to know about like what it's like to fight their fights in a way because it's pretty sexist in our field. And these people are succeeding. Although I'll say Dr. Ketchy, the third one, she's not in data science, but she's in science, right? Like she's doing research. So she's running the same buzz saws that we run into. Maybe not exactly the same that you might, if you run like a data warehouse or something like what I've run into or probably what Dr. Debru has run into with her AI projects. But there's sort of the special sets of problems that you get when you get, when you're a woman and you get a certain level of power, like when you're sort of lower, they, you know, people dismiss you and stuff. But if you get power, like men start to attack and not all of them obviously, there's always just, it's a minority, but it's a loud minority. And you know, the majority doesn't stop them. That's the problem. The problem isn't so much the men that just say things and meetings or just do things, just take stuff from you or whatever they do. What matters is the men who stand around and don't stop it. They don't stand up and say, hey, cut it out. What are you doing? You know, and so that's been my problem in life. It's not so much the aggressor guys, but everybody else who just stands there stupidly and I have to stand up for myself, you know? So whoever's watching this, I wanna educate you that this happens to women. And whether you're a woman or a man or you're, you know, not choosing a gender, you're not a buyer or whatever, it doesn't really matter. It doesn't matter what your gender is. We need to stand up for people when they're being treated like this. And so now I'm gonna start. And luckily, the first person I'm gonna talk about here, I don't know what, I don't, none of the stuff out there makes me think that she struggled, I'm sure she has, we all have. But she's just like a rock star. Everything looks perfect, right? And her name is Dr. Kimberly Sellers. I don't know her personally. I connected with her on LinkedIn and I've been looking at her research basically. So she does this thing that's impossible for me, which is math. She's a professor of mathematics and statistics with her specialty in statistics. But one thing that really first attracted me to her research is that she's an author of an R package. And that's kind of one of my bucket list things is, I'd like to author an R package, I don't know about what. So obviously I'm thinking about the wrong way. Dr. Sellers' research is on fitting Conway Maxwell Poisson regression models. Now, I don't know a lot about that because I usually don't fit that. But I know with Poisson and it's rates. So she co-authored this R package for that. Her specialty is studying dispersion. So she's really into the math and theoretical statistics side of it. Whereas I'm in more of the applied statistics. But what's important is you see that she's at Georgetown University. And one of the things I wanted to show you, she has a personal webpage. Let's just open this when I click on it. Yeah, so this is her personal webpage, well, personal professional. And look at how gorgeous it is, right? Like she didn't have to do that, but she did. And one of the big things on there is teaching. So one of the things, one of the take home messages I want everybody to get from this is that before you choose to go study for your PhD or whatever, data science, you can learn a lot about the professors at these colleges. And right now, if I was like a person thinking about studying mathematics and statistics and I was kind of excited about Conway Maxwell Poisson regression models or I thought I could get excited about it, I would go here and try to work with her. So I also link you to an old QA where it's old, it's like from 2015, but what I liked about it is she sort of explains her teaching philosophy and other things and this is her R package. Of course, I don't actually know how to fit this distribution. So I'll have to learn a little bit more before I do that. Oh, Daniel, here you are. I just looked at the chat here, so let me share here. Good to see you and attend today's session. Let me show it. Well, thank you, Daniel. Daniel is my SAS superstar on LinkedIn. And let me know if you have any special requests topic because I hardly get to talk to you and I'd love to tell you anything that you're interested in. All right. So then let me go back to our data science update. Let's see here. So the next female scientist on our list is Dr. Jebrew. And Dr. Timnit Jebrew is now the founding and executive director of the Distributed Artificial Intelligence Research Institute or also it's a dare like that. That's what she is now. Now, I don't know her personally, but which is ironic because she grew up near where I live. But I don't know her personally. And actually I guess Wired Magazine, how I learned about her research is actually from this really good profile in Wired Magazine here, which I keep around because I was very inspired by reading it. There's a lot going on with her, okay? So first we have to go back to think about artificial intense for a second, okay? So let's just remember what that is. It's a model, right? It's an equation. And like it's like a regression equation, but it's got more slopes and stuff, but basically it's a predictive model. So how do you get an AI model? Well, you give it a whole boatload of data and you tell it what the dependent variable is and you try to get it to fit like the best predictive model of that dependent variable, okay? So you're following me. So if you're in a racist society that disproportionately jails black people, then of course, if you take data about people who are jailed, you're gonna come up with people who are black, people who are poor, you know, all the people that are being oppressed in our correctional system. So it's not predictive necessarily. It's way reflective. It's telling us what our values are, like who belongs in prison, okay? Well, that's a problem because if it reflects back our values and it doesn't say people who commit crimes and are violent and dangerous belong in prison. It says people who are black and poor belong in prison. We screwed up. We screwed up our prison system, okay? And so the problem with taking an AI model, making it, and have fitting it to the dependent variable is you have to think about what the independent variables actually mean. Everybody hates that part, right? You have to actually think about them. One of the big problems, and I'm not an AI expert, one of the big problems in AI models is people will fit something that is not allowed to be in the model. Like they'll even do this in regression where let's say like I have this guy who had all these mice and he was trying to predict their weight at the end of the study. And he wanted to adjust for something else at the end of the study. And I'm like, you can't do that because you wouldn't know that at the beginning of the study if you're trying to predict it, right? And he could not wrap his mind around that. So those are, that's a common AI mistake to put something you couldn't have known in to predict something that happens later, okay? So that's its own problem. But otherwise, like I've heard of where they try to do classification of images and then they build an AI model and images of stuff in the kitchen is then correlated to women because we're sexist and we relate women in kitchen. And so this is the problem with AI. Like that's the problem in general with AI is that it reflects our biases more than like some magic predictive thing. So that's a problem if you're trying to do something like sell stuff and everybody's trying to sell stuff. So Google got concerned about the problem of that, of this bias problem. You can do stuff about it. Like obviously this problem exists in people research it. So there's stuff you can do about it, right? But they want to, obviously if it's Google they want to solve it. So they put together a team that was led by Dr. Jebrew that in a way it's kind of ethics, right? Because you're trying to not just, you know keep black, poor people in jail because you're racist and you set up this correctional system that now produces AI models that say that, you know what I mean? Like that cycle is bad. What you're really trying to do is something practical like predicting who needs help that you can go help them. You know, not, you see what I'm saying like something useful for capitalism or something. So probably Dr. Jebrew and I are similar in that we're researchers and we do hypothesis driven analysis. And so when I worked at the army at Natic Labs I was doing that too. And I had the same exact experience that she did that was actually written about in the Wired article. And that's where she made a peer reviewed article and went to go submit it and her bosses won't let her submit it. And the article wasn't even like that controversial. And that literally happened to me. I had an article that was benchmarking how fast it was to load data sets in SQL server versus PC SAS cause I don't know why we're using PC SAS but I was trying to show them how like lagged it is. Nobody believed me. So I just did the study. It's not a controversial study but I could never publish it. It was a beautiful study. And so I know how she feels. Well, this is what I did. When I left the army I just started my own business, right? And she almost did the same thing only what she did was a little more grand, I think. And that is that she, I'm sharing my screen again here is that she set up, is this working? Yeah. She set up this distributed AI Institute, Research Institute. So she, you can go there. There's not much on the page. I keep going back there to see what's going on because of course wouldn't it be cool to work there, right? Or do something for them. She's gotten her team together, I think. But you definitely want to watch her research because you know if it was suppressed, it's got to be good. All right. So let's move on to the last researcher I want to highlight. This is my close friend, Ketchy. I just love Ketchy. I don't know if she's on this live stream. Did I invite her to live stream? I don't even know if she's on LinkedIn but she used to live in Boston. That's where I live and that's how I met her. We used to work together and then now she's really taking off. She just received full professor tenure Central Michigan University. She's the head of their nursing program. She's the director. I looked it up to make sure it was director of nursing. So if anybody's out there and you want to go, especially if you have an associate's degree in nursing and you want your BSN, definitely contact Ketchy because that's what she does. And so she's mainly a nurse educator and administrator. She used to, I guess, do patient care but by the time I met her she was the nurse educator and administrator. And she really, really, really understands nursing education which is super complex, okay? But I've known her like kind of a long time now maybe like 10 years even. So I've watched her research sort of evolve. It started more towards where she was focusing on like improving nursing education and what it would take to improve nursing education. And that sort of segwayed into the fact that nursing education is just so racist. It's like if you're not white you usually have a terrible experience in nursing education. And I mean that, so you can't just look away if that's the problem. So it sort of evolved towards institutionalized racism in nursing and she was featured in, her research was featured in this USA Today article how nurses of color fight the rampant racism playing their field, that's what I linked to here. And I encourage you to read that. And I also encourage you to go to her Google scholar page because she keeps that up really nicely. And it's got her research in it. And it's really interesting because you know I'm in public health. And what my concern is, is that is healthcare quality. Those of you who look at my other videos and go to my blog, you'll see I do a lot with healthcare quality. I did a lot with a big client in Qatar, their healthcare quality department. I was working with them to do research to improve healthcare quality. Well, if you have racism, it's gonna ruin any effort at healthcare quality that you make. And now that the COVID-19 pandemic happened, in the US we can really see the effects of institutionalized racism in the health system. You know, so it's almost like public health has to turn its attention to institutionalized racism just the way nursing education has to. Because from the nursing side, what has happened is nurses have been told that they're the solution to the US healthcare system being fragmented and not efficient and whatever. Why, why are they the solution? You know, I don't know, there's always dumping on women. You know what I'm saying? But that's been promoted for so long. So Dr. Ketchy's kind of like, well, if we're the solution, we need to get our act together. We can't be racist. We can't have racist education and racist stuff going on, like at least that. So I really like it in a way because we're from totally different fields but we're kind of going for the same issues, you know, that we, you know, just want quality healthcare and quality education, you know, me in public health and data science and her nursing, but you know what I mean? It's like pretty important. Well, now I keep checking the chat here. So if anybody's got any questions, go ahead and ask them. On any topic, I'm just giving you the update of what I've been up to because I was so busy. I wasn't really doing anything with those researchers but I was looking up them and following them. And so that was part of what I was doing for those, for women's history month this month and black history month for last month. Oh, and by the way, I'm on the board of directors for central Boston elder services, which is they're an elder, they're a nonprofit providing home care services to African-American elders in Boston. They have, there's a nursing home but they also provide other services, you know, in-home services. They're really nice member of our community and they did a lovely video. They wouldn't let me embed it here. It's just wonderful. I put a screenshot from it up here for black history month. It just makes you feel so good when you watch it. And also please consider donating to central Boston. I'll tell you where your donation goes. They literally go out and buy like walkers for these people or pay for their dentures or when it gets really hot, like maybe they'll buy them an air conditioner. Like they need money like that they can just spend on their clients that's not regulated by the state or anything like that. So, you know, if you give money, like if you gave like 50 bucks, probably somebody's gonna use that as a real like object in their home that's gonna change their lives, you know, like they'll be able to see cause they have classes. So if you've ever are like, oh, I'm worried this will go to administration. If you donate to this link, no, it's gonna go to buy that stuff for those people. And then let's see here. Now I'll move on to the, my personal update. So this is something I'm very excited about that I've been doing with my forever intern, faithful forever intern, Natasha Dukacz. And she, I think you probably wanna go to her site. She's Ukrainian and she's over there trying to help everybody. And she also has like a charity where you can give, where they can give cash to people who are fleeing. So you should go, I think this is her LinkedIn site. So go look at that. But before all that started, you know, months ago, few months ago, we were just in Boston together. She lives here with me and not with me with me, but nearby and we're kind of neighbors. And so we can meet, we can do collaborate really easily. So we've been collaborating on sort of a series of projects to do with dashboard design and functionality. You know, I have a background in design. I have an undergraduate degree in fashion design, but I learned graphic design. I've learned a lot about design and like user interface design. And I'm just terrible. I cannot program front end to save my life. So Natasha, on the other hand, she's kind of, you know how I was saying like, Dr. Sellers is so amazingly smart that she's so mathematical, whatever. Well, Natasha's also just dashingly smart. I don't know how she's figured all this stuff out. Like I don't even know why I call her my intern. She's like learned all of it on her own. But anyway, we've had this sort of line of where we're working on dashboard design. And she's actually building these dashboards in different R packages. And she's fearless. She's just fearless. She'll try anything to get things to work. And she's amazing. So what our latest stint was is she found this dashboard online. I encourage you to visit it. I couldn't make head and her tail of it, as I said on here. And I just kind of ignored it, but she really wanted me to look at it. It really infuriated her and then started to make me mad too. Because what it was supposed to be is a dashboard that helps people like me and Natasha figure out what hospital go to that has the least rates of infection. And it doesn't work for that. Like I can't, I'm an epidemiologist. I can't tell what this dashboard is saying. I can't tell what the date is from. I just like moved aside. I looked away and she's like, no, you have to look at the ugliness. And what I found is it's put up by one of our offices. We have a big government in Massachusetts because we're in Texas, we get a lot of money. So I don't have a lot of money, but our government does. They had enough to put up the dashboard. And so it made Natasha so mad. And she started asking around. A lot of people pointed out that the dashboard data in there was probably from the National Health Safety Network, which is not really that good of data. So the data is probably not that good. That's what I thought. That's what everybody, but she's just fearless. Like I said, she said, well, let's scrape the data out. Cause I think I'm doing that right. National Health Safety Network, NHSN data. That's not private or anything. So it's just a matter of requesting it and getting it and putting it in your dashboard or whatever you're going to do with it. So she scraped it out. She just scraped it out. Like you can read our book chapter. We published this book chapter and it explains how she scraped it. It's got the packages she used in procedure, roughly the high level one. And she's got a GitHub in there so you can look at what she did and the links to the dashboards. But if even the here you can see, it's just gorgeous. She redid the dashboard. But what the whole chapter is about is how frustrating we were that the reason why they put up this dashboard here was due to open government data standards, but we couldn't use it. Our taxpayer money basically went to this dashboard that hid the data from us. And so we were mad about that. So that's what this is actually about is about how to evaluate these open government data dashboards like it provides a system you can use. So that's the heady part of it, the sexy part of it that I should do. The heady philosophical part was mine. And we just submitted an abstract to the AMIA meeting. It's going to be in DC in November. It used to always be in DC and I always went. Then it started being like San Francisco and stuff. And I just didn't want to go. I love going to DC because it's especially now that I live on the East coast it's so easy to go to. It's like going to New York, but no, I didn't want to go all the way to California or wherever they, I think they held in Orlando once and I had it with Florida. I lived there for five years. I was just done with Florida. So it's a good vacation, Florida's, but I just feel like it. But now DC that I like it. So hopefully they'll accept we applied for like a system demonstration so we can demonstrate our dashboards for you. And hopefully they'll accept it. And then you can come to the meeting and we will show it to you. But if you, oh, hello, we have a chat. Hello, ma'am. I'm one of your students. I learned a lot from you, but I know that data science is like an ocean. As much as you come closer, you will go deeper. Can you please share some website or application other than Kegel where beginners can practice and make, I guess this is long. So let me see. Can make models. Okay. So welcome to fun. So I'm gonna reinterpret your message, your, what you said in a different way. I'm gonna say like what I think you're asking for is where do I go to get ideas for a portfolio project? That's what I'm gonna call as a portfolio project. Get data for that portfolio project. And then get guidance on how to actually create a project. Okay. Well, let's first talk about the fact that it's really hard to even do that if you're a professional citizen. Okay. So let's say that like Starbucks came to me and they said to me, we don't know why certain coffee is not selling. All the other ones are selling a lot, but this one isn't. Okay. I'm welcome to my portfolio project, right? Like let's say I sign a contract with Starbucks. They'll let me have their data, but what data do I want? Like how do, I mean, there's tons of data at Starbucks. There's transaction data, there's internal data. There's, you know, like what data would I have? And how would I phrase these questions? Like when did this coffee become unpopular? When did it even start being sold? You know, like I'd have to sort of formulate these questions. And then whatever I ended up doing, like I'd have to go from their sort of hazy question, like why isn't our coffee selling? That's not a very clear question, but that's what clients say, you know, what's wrong? You know, why isn't my database working? Why aren't people coming to my website, you know, like that? They don't ask like hypotheses and they're all, you know, H0, H1, you know. So you're gonna have to do that yourself. And you're gonna have to figure out how to do that. And you're gonna have to figure out how to pick from their data, even if you're at Starbucks or whatever, you know, so let's say that you're a data scientist and you say, okay, I know how to do it. I know how to re-ask your questions in a way I can answer them and I know what data to get. Well, then the next step is you've got to actually do the project and communicate to them the answer. Like you've got to get the answer. Like here's why your coffee is not selling is it's too bitter. Nobody likes it because it's too bitter. Or it's more expensive than the other copies or it's oilier and they don't like the oil, it's on the top. Like I don't know, you have to find the answer for Starbucks, right? And you got to do that, however. So now that I've reframed sort of this issue too fun, you'll probably see that maybe it's not like you should be able to just do a portfolio project. It's like maybe you need someone to walk you through it, right? So what happens is if you have a lot of money and a lot of time, maybe not a lot of money, a lot of time, but you have money and time. What you can do is you can pay somebody like me or other people, there are other people out there that specially does different things. And what we will do is guide you from the beginning to the end of this portfolio project. Now that, what I'm describing that right there is the sort of like top shelf data science experience. Is if a tutor like me or a mentor like me who's got experience and does this all the time sits down with you from the beginning to the end and shows you how to do it. But even if that happened, that's kind of like a person getting a PhD. It's the first time they've ever done the whole project from beginning to end. They still, they don't know how to do it. Like they're gonna need help the next time they have to do it independently even though they have their PhD even though you just did that. So what I would say is that on the way to doing that, there are these things that you can do where you see how someone else did a portfolio project and you can kind of walk in their footsteps. Now when people say, what about Kegel? I say that is exactly what Kegel is. Kegel is something where they define the problem to some degree, they got you some data sets and now you just have to finish. Maybe starting with Kegel and learning how to do that is a good place. But that's not where you end. That's where you start, with these putting some pieces together. Now what I tell learners is they should try to find a topic they like and try to find data sets on that topic no matter what they are. And it's easier now because we have social media. You can get data sets from social media about people talking about a topic. That you get, you find some data about a topic and you find some problem with the topic which if it's a topic you're familiar with you already know the problem. Like I met a guy who was into analyzing basketball and he was explained to me this problem people have when they're trying to bet on basketball games that they wanna know certain metrics and whatever. And so he already knew kind of what people who bet on basketball worry about and what they wanna know. I didn't even understand what he was talking about. But he's perfect. I didn't understand what he was talking about. But I could guide that guy with his data to produce a really nice portfolio project that if somebody draft Kings or somebody wants to hire him into these gambling places sports gambling places, he'd look really good to them. And that's sort of the way I frame it. As I say, think of the end. If you wanna work at like a marketing analytics place do analytics about somebody's marketing. Go say, oh I saw Zara has a new clothing marketing campaign. Let's see if people are talking about it on Instagram. You know what I mean? Like you can come up with questions that you can answer with public data. It's just not that easy to do when you're just swimming out there and you're new. Like how do I know what question to ask? How do I know what data is gonna answer it? Yeah, that's really hard. And again, in the beginning what I would say is see what other people do. Like read a lot of other like good case studies where somebody good. Like I'm always telling people to do portfolio projects and I'm always telling them to post them. Well, look at other people's portfolio projects. Don't copy them. Cause that's, I mean, you can copy them if you wanna just practice, like see what this person did and have their experience. But don't copy, you know, like if somebody asks a question like, you know, what about Zara's new campaign on Instagram? You know, then, you know, for shoes. You know, don't ask almost the same question for like jackets or something like that. You know what I mean? Instead ask for, you know, saxet avenue or something like that. You know what I mean? Like make the change different enough so that you're gonna have to change your research methods. You're gonna have to change everything. Now, what I often do at this point is I pitch my LinkedIn learning courses because that's what you're supposed to do but the real truth is I've not found any courses that help people design studies with big data. There are a lot of courses out there that can help you program and learn how to program. And there's a lot of courses out there that can teach you like how to use software in data science and also, you know, how to make nice presentations but it's, there's nothing really that teaches you how to ask a question and design a study with the data if you've got that you can use to answer that question which is what we're all trying to do in data science. So in my LinkedIn learning courses there's designing big data healthcare studies part one and part two. And it's not related to any software. It's just a way you think about formulating questions and answering them in data science. Like one of the things I point out to people is you're gonna have data systems that are constantly making production data. Like, oh, somebody went to the emergency room. Here's the ambulance. You know, they're updating their systems and you know, the emergency room's updating their systems but if you're trying to do a study for an ambulance group to try and like minimize wait time between dropping off people at the ER and picking them up and stuff then you're gonna have to know like how to design a study that's gonna answer exactly that question. You know, and the only way you can really do that is if you get some stable data that's not changing every two seconds. So you have to learn how to actually take an extract from the production data I was describing. You have to learn like I imagine production data to be like the stream. You know, I'm from Minnesota. So there's like the stream going by a lot. There's a stream. And then you just wanna scoop up a little data just a scoop up the sample of data and study it. You know, just like they would just like an environmentalist would study a stream is that and you study that static data that you get from that ER ambulance or whatever. And you ask your research questions of it and you answer it. So let's say our first research question was well, how long does it take to turn around the ambulance after they drop off a person at the ER? How long does it take? You know, that's a descriptive question. Well, that's a hard question. You know, you're gonna have to get some data. You're gonna have to probably do some count queries and some sums and whatever and answer that. Because then the next question is, is that too high? Or maybe it usually is too high. You know, how do we reduce it? That's the next research question and that might be the next set of data you have to get. So you have to get good at doing that. If you're a data scientist, no one else is gonna know what to do. People are gonna know what data is there because you ask them about their business. You know, what do you do when you drop people off at the ER? You know, and you're gonna see they update these things and so there's data. But they're not gonna know what to do with it. Like that's all you. So you said another application other than the Kegel where beginners can practice and make models. I would tell you, there's this one place where I know you can get a lot of data. Like there are places where you can get data. There's, I'll give you a link to it. R-A-P-S. This is socio, this is sort of health data. When our government in the US collects a lot of data that they don't really need anymore, but they want people to be able to analyze it or other places do. They often put it in this repository. It's called ICPSR. It's at University of Michigan and Arbor, I believe. And so that's one example of a place. But it really depends on what. So first, one thing you can do is go find data or figure out data that you can use for a project. But the problem is if you don't have a topic, you're not really sure what you're looking for. So it's better to pick a topic, like am I doing marketing? Am I doing health? What am I doing? And then try to find the data and learn about it and then learn how to ask a question and answer it. So if you took my LinkedIn learning courses, for example, so you have a study design background, then you picked a topic and then you wouldn't did that. You found maybe some data. Let's say you'd picked a healthcare topic and you found some data on that website I just put there and you were able to analyze it, apply a study design, ask a reasonable question, apply a study design, analyze the data and answer it in some sort of blog post or white paper or peer reviewed article, then you're a data scientist, welcome. And you're done, you're there. And you can go get a job and tell people, you know how to do that, right? And you can do it again and again and get better at it. But that's basically kind of, there's not really like one website. And also what you'll find is, as much as I put that link there as an example, you'll find data sets from like the 70s and 80s there. You'll find data sets that you really can't get much out of, like you probably wouldn't be able to do much with. But they might be, like I always tell people, try to do something relevant for now, because people will say, oh, that's really interesting. Actually, let me share my screen because I'm gonna show you something that's related that was in my update, which is that a long time ago, I wanted to show people an example of this advice I'm giving. And so I did this thing where I analyzed data from our local casinos, which was government data, because I like to go to the casino and my friend and I were so funny, we're even going today to the casino. We always complain about the restaurants. And then we go anyway, we just hope that we can find something good to eat. And we're like, that makes no sense. If we're complaining about the restaurants, certainly other customers are. And so I have this casino blogger guy, Robin Oben. He's so wonderful. He likes to travel around and tell us in New England where the best casino experiences are. You know, where is the funnest, best restaurants and just the best experience. And the name of his blog is New England Time Gambling or any Time Gambling. I never really said it out loud, actually. But anyway, when I did that casino, I did a casino comparison and I gave it to him and he put it on his blog and he just did an update. Like we talked, you know, it was before the pandemic. So we all, we talked basically, what I talked to Binbin, that's what his name is about, is hypotheses. What are our next analyses? Like one of the things that Binbin said after I posted that the first time was that he said, Monica, you know, Massachusetts is a small state and we're mainly competing with Foxwoods, which is this big casino in Connecticut, which is next door and state next door. Why aren't you analyzing that Connecticut data and vis-a-vis this Massachusetts data? Because that's been the big issue of Massachusetts delayed opening casinos because they're like, well, there's Foxwoods already there. But people in Massachusetts were like, well, we could make taxpayer money or whatever. Why don't we have our own casinos? And so now there's competition. The question was, should we even do it or would there be too many casinos? But now I'm not worried about that. I'm worried about them having good restaurants. So of course my friend and I, she's a fintech data analyst and she's an accounting and stuff. So together we wanted to work on, do a contract with these casinos. Obviously they're bleeding money. We end up giving them money each time we go. But we haven't been lucky enough to get their attention. So if you talk to the Massachusetts Gaming Commission or on core casino, tell them Monica and Josie will do the best analysis for you, give you the best advice. Because actually I'd love to do it because we've been there so much. And I look around like one of the things that she's noticed is that there's not a lot of, like, well, I think it's changed, but in the beginning there weren't a lot of people of color there. And I was like, yeah, there's none. And she's like, well, there's nothing here to attract us. You know? And I was like, yeah. Because I'm like, my dad's from India. So I'm half Indian and I'm vegetarian. And, you know, the Asian food they have at this casino is not for Indians, you know? It's very greasy and it's not vegetarian. And so I'm just like, okay, they have like three Asian restaurants that are all not good for Indians. Come on, you know? It's like some of it seems really obvious. But anyway, you know, data science, that's the thing with data science. I'm like, don't do the data if you can just answer the question, obviously. But anyway, so that was an example of sort of a portfolio project that you would do if you were trying to get hired by the casinos, although it hasn't worked for me. I've tried that hard, but. And oh, I also wanted to go over some of the other things on here that I've been doing too, which is we went over Natasha's dashboard. I've been doing some work with the R survey package with the BRFSS data doing weighted analyses. And I think I'm gonna publish something about it. But if you want something right away for some reason, like if somebody really needs the help, I'll let me know and I'll do it sooner than later. Then another thing I've been dealing with is predatory journals. So I just wanna tell everybody how you know a journal is predatory is it comes from like a printer, a publisher that is a predatory publisher. So legitimate science publishers include like Springer, Elsevier, ScienceDirect, these are all familiar to you. We're like Wiley. The predatory ones are not those. There's a list called Bell's List. I encourage you to look it up. Maybe I'll put it, sometimes I put a nice description on these live streams. So if I get around to doing that, I'll put it on. And so what the problem is, is my colleagues get these emails from them. Let's say, oh, do you wanna publish with us? And they think that it's a good idea. And I'm like, no, it's not, like some of the times their names are really similar to legitimate journals. So I've been doing a lot of that, like running interference there. Then also I've been working with my lipidema group. We're working on a paper. And when it comes, I'll let you know. But we, someone joined our group who knows a lot about lab values. So I was like very grateful because that's my weak point. And then also just like if you ever, if anybody's watching this after it's recorded, or we're watching it now or whatever, and you ever have questions for me or you wanna talk to me about data science, just contact me for a free consultation by email or LinkedIn or however, and we'll set up something on Skype or Zoom and I'll talk to you about your thing. All right, and then I just wanted to let you guys know that there's this telehealth session next week, this Parks Associates, I think it's kind of expensive if you pay, but I think I got lucky and got a free ticket a couple times. It's really cool if you can go to these online sessions, they're gonna try to do them in person. I'd love to go to one in person one. They get people who are in, it's not just healthcare innovation, but just technological innovation. I tend to come to the healthcare ones where they're talking about their new innovation and sometimes these things are pretty early on, like they're early on startups. And some of them are really kind of cute and stuff, like you kind of know it's not ready for prime time, but you're gonna feel like, it's kind of like seeing a local band before they become famous internationally. That's the way you wanna kind of see these talks. So I enjoy going to them. Then the other thing I put on here is about Timon Vinkie, I hope I'm pronouncing your name right. He's a business coach and I need business coaching. Just so you don't like some of you think, I need a data science mentor. Well, if you do, call me, right? I'll hook you up. And so businesses sometimes feel that way too, but the problem is you're just not really sure like who's gonna help you. And I would say, this is me personally, it's worse in business than in data science. The people in data science, we're not perfect, but we tend to be more on the honest side than people in business, unfortunately. There's a lot of like business coaches that I don't think they even know what they're talking about because if you're in analytics like I am, if you're in informatics, you take some business course or whatever, and you're just like, this is BS. And so what I liked about Timon's approach is it's sort of like mine. And so where he was trying to explain what he does, like if you pay a business coach, what do they do? And I even wrote a blog post that says, what does a paid data science mentor do? Just so people can set reasonable expectations. Well, Timon's holding this free business boost program, but it's like, he invests a lot of time in it. So you only have so many people in it. So you wanna sign up for it. And it's a few sessions and it's to help you like understand, I think, the best I understand it, how to get business coaching, what you can get from it. He's been successful in bigger businesses and smaller businesses. So I'm like, well, that's, somebody already is like, you see how many publications I have, how many things. You know, if you come to me and you need to do peer review publication, we won't have any problem because I've done it so many times. So that's why I was like, okay, in business, when you're kind of worried about getting coaching, he seemed like he had a really good program for me to try it out and see if we click because that's another thing is, whether you're doing data science coaching or business coaching, you really have to click with the other person. And there have been times where I just said to customers, we just are not a good match. We're not, if that happens, I don't charge them. I just say, we're not, we can't really work together. We're not, our styles just don't work. And so that's another reason why, well, that's part of the reason why I have all these LinkedIn learning stuff and I have the stuff on YouTube is, I've got kind of like a strong personality, but they always say that about women. They said that about Hillary Clinton. They always say, people have always called me abrasive, but I think that's sexist. But anyway, if people have problem with my personality, then they don't have to hire me. And the fact that I have all of this information out there about what I'm like makes that an easy decision for people, hopefully. So that was mainly my data science update. What like I said, I'd been, what I really have been doing a lot of is, I have a colleague and I already just love working with her because she's like an epidemiologist and she's those dentistry. And we've been really looking at oral health and the BRFSS and issues with oral healthcare in the US and we put together some papers, but I'll cut you to the chase. It's a big problem if you don't have access to dental care in your population. Even if you have access to medical care, because a lot of people don't realize this, but diabetes that can cause your teeth to fall out. Like having chronic illnesses can destroy your teeth. And so if you don't control your diabetes, if you don't control it, you could destroy your teeth. And so actually oral health practitioners really need to be integrated and working together with medical practitioners, like if you have a diabetic patient. And to some degree, this is sort of normal in other countries. Like I have a lot of customers in Saudi Arabia, they have a big community health dental infrastructure. They have so many general practitioners in dental and they are well aware of the connection. They have a lot, unfortunately, a lot of chronic disease in Saudi Arabia, high rates of diabetes, high rates of now I'm finding somebody's telling me liver disease. If you look it up, there's some papers on it. And these people are a total risk for their teeth falling out, but they have this whole oral health care system around them that's supporting them. It's just that Saudi Arabia has its own public health problems because people are different everywhere. Like here in the US, we had this anti-vax stuff going on. In Saudi, it's different. Like a lot of people have diabetes, they don't wanna get tested because then they'll get the label of diabetes and then they have to go deal with it. But isn't that better than just not knowing you have it and being sick and not knowing what you're doing to your body? So that's one of the big messages in Saudi Arabia. It's just go get tested, see if you've got diabetes because if you got it, Saudi Arabia's there to help you manage it and feel better. If you're in the US, good luck. I don't know for your help you manage it. We charge a million dollars for an insulin. I don't know what's happening in the US health care system. If you've been watching my live streams and following me, I did a project for Yale University. I did some lectures for them. And one of the questions that I need to shine some light on was what's the future of our health care system? I just have no idea in the US. I really just don't know where it's going. It's crashed. And other things are taking precedent over our health care system. And so nobody's really cleaning it up and trying to figure out how to get it to work. People still come to work. There's still all these hospitals around me. I still get charged my insurance fee. I was just thinking today, I don't know if I've seen my practitioner. I don't even know when the last time I had my mammogram was. I'm paying someone. But I don't know if I'm getting any care. I mean, I can afford dental insurance. So I'm getting dental care. That's another finding from one of our papers, I guess. It's just, you need to give people dental insurance in the US or they're not gonna be able to access oral health care. All right, well, thank you to Daniel and to Fawn and whoever else showed up today at my data science chat update. And don't forget to like this if you liked this video and subscribe if you want to have more opportunities to chat and more videos like this. And also there's other videos. There's videos of me showing you how to use SESS, how to use R, how to get a job, how to do public health. I don't know. I'm trying to build up my YouTube channel more. I already have a lot of good videos but some of them are a little older. So I'm trying to refresh things. And like with this update, put some new material on. All right, well, thanks very much for showing up on this Friday. I hope if you're watching this in real time that you have a good weekend and I'll see you on the next live stream. Bye-bye.