They know when our flights are and what we are coming with, and I've requested an attendant to walk us through security if they have one available, and I think that'll make the process easier.
To sign us up for that all I did was call TSA and I'll put some info down in the description. For the items that are typically banned, like the needles, the liquids, exceptions are made for medically required supplies. And, they're probably going to search our things thoroughly check this website. They can't open anything like they can't open my formula, they can't open the needles, but they'll search it. They're probably gonna do some extra screening on us, but exceptions are made because it's medically required, we need it. For Harlow, I've actually flown with her before, and I have a video on it here. But, it was pretty simple going through security with a service dog. I went through the scanner first. Harlow held a sit-stay, and then I called her through, and they patted her down, that was it. For the plane ride we sit in bulkhead, which is the front row where you have more legroom and she just held a down stay at my feet and she basically slept the whole time. Something else I have that I don't necessarily need, but it might be helpful, is a letter from my doctor stating that I am on IV therapies, these are my various IV prescriptions, I need the use of needles and, basically, like I need all of the supplies for my medical conditions, please allow my patient to travel with everything she needs. Just in case we run into issues with security or something else, I'm covered with an official letter on an official doctor letterhead. Well, y'all, out of the house; we are going on that Target run I needed. My mom is here, and I have a girls' afternoon. Right, Mom? - Yeah, out with my baby boo-boo. - And Harlow. - [Mom] And Harlow. - We're looking for - [Mom] Where are you? - I'm over here. I'm the one with the dog. (laughs) (customers talking) - That's not bad. I'm looking for a jacket or a hoodie for like traveling, that's nice. - [Mom] What is S? - Let me see. I can't decide if I like the color or not so my mom said to hold on to it for a bit, and then we'll see. What do you think, Mom, do you like the color? - [Mom] I do. - You do? - [Mom] Yeah, I think it's a really nice color. - Okay, maybe? Okay, this is what I need to keep my infusion bags cool on the flight, so it's an insulated sack, basically it's called the lunch sack. And, I don't need it to, like, essentially close; but as long as I can seal it enough around the infusion bags and put in some ice packs, it'll be good. And this is, like it gets really big, I just need to fit two for my trip. So this is perfect and it's only $3.99. So I realize this is basically the same color as the shirt I'm wearing. I actually like the color, (talking via PA system) but it's kind of big on me. But, my mom said it's good that then.
0 Comments
So, in these organizers are organizers, organizers within the organizers.
And, it has my IV medications. So, what I'm gonna do is in the two empty slots, I'm gonna put in some of my port care stuff like my skin prep, my Biopatches, and claves. And then, I'm gonna fill the other one with the swab caps that go on the end of my port here when not in use. Then, I'm going to take the top here. (tops clicking shut) And, this is how I'm going to pack my IV medications and some port supplies. This stuff is gonna stay in Orlando because I'm packing extra for our trip. And, that's how it's gonna be, I'm probably gonna put some paper towels or toilet paper to fill the gaps so the little vials don't rattle around. And, then this suitcase I bought specifically for this trip from Target, it's hard shell and spinner so the wheels go every which way. And, of course, it's pink. All right, what else we got? So, call me a little overcautious, but I'm bringing an extra feeding tube button in case my current one malfunctions while we're over there. I don't want to have to make it a hospital trip, I can just easily replace it so I'm bringing an extra one. And, then I've got some feeding tube supplies like gauze and tubing pads, extension tubing, my feed bags. You'll notice I only have two bottles of formula. That's because Kate Farms is being absolutely amazing, and they're sending a box of my formula to where we're going, so I don't have to worry about packing like 10 plus bottles. They are incredible company. And, then I've got my needles for my IV meds, my large tegaderm I use to shower while my port is accessed, and then I've got supplies to access my port. I'm actually gonna access it the day before we leave so it's good for a week. But, in case it gets wet, or I need to access and re-access, I have the supplies. It's good to have it, just in case. Saline flushes, heparin, batteries for my pump, alcohol swabs, of course. And, then I'll be taking a Lactated Ringer and a saline bag and the stuff for my banana bags. And, of course, this will be packed at the last moment because it's gotta be refrigerated. To keep those cold for the flight, I'm planning to buy little insulated packs and also use ice packs. Actually need to go to Target before our trip; but not complaining about a Target run. And there we go, everything fits nicely. But of course like the infusion bags and my pump 'cause I still gotta use it, that'll be packed at a later time. Now, let's talk about security. Security may seem a little tricky with wheelchairs, service dogs, needles, liquid meds, fluids, whatever else. I have signed us up for something called TSA Cares, which is a program for people with special needs who are going through the security checkpoint. - For the airplane, I'm packing things differently.
And there we go everything fits nicely. Now let's talk about security. (energetic upbeat music) Good Sunday morning everyone, Harlow is just made herself comfortable here with the headrest. But, anyways, two updates for y'all. IV granisetron definitely helped me last night. There was still some vomiting but no dry heaving, and I slept better, so I'm just going to wait and see how the next few days go. Also, my new feeding tube has arrived, hopefully this one is the right size and it'll help with the pain I've been having. But, Judd wants to be there when I change it for the first time by myself, I want him there to. He's working today so we are going to do this tomorrow. Today, I am packing my medical supplies for our upcoming airplane trip. So, this is how my coming week looks. On Monday night, Judd's mom Ann, is coming so she can take me to my ketamine joint injections Tuesday morning, and from there she's going to take me to Paul and Jenese's house, Judd has to stay here to work. Then on Friday, Judd, Paul, Jenese, and I, and the pooches are leaving for our trip. We're coming back Monday, we've decided to keep where we're going and what we're doing a surprise. If you're wondering why we're doing this when our health isn't the best, well, once we get there, you'll understand, it's a once-in-a-lifetime thing, there's a lot of love involved with this trip and it's actually been planned for like six months. And, then Tuesday, Judd and I come back home. So, I'm actually gonna be gone for a week. For Orlando, this is my new awesome travel organization stuff and if you want to see the vlog where I go into detail about these things, check out this video. But, this has everything I need for Orlando, I just have to bring some more formula 'cause I'm gonna be there a little longer. But it's got my IV therapy, port stuff, feeding tube stuff, I'm set. But, for the airplane I'm packing things differently. So, for medical supplies, I actually get a complimentary checked bag, same goes for service dog equipment. But our trip is not that long and I'm not comfortable checking my medical supplies because if the bag is lost or it's handled roughly and something is damaged, I'm basically screwed because there's no way to replace this stuff, so I'm actually bringing a carry-on just for my medical supplies, and I've called, verified that's totally okay, and this is how I'm going to pack it. Sorry, Harlow, I need your headrest here. You look so cute, though, sorry. So, for now, I'm just going to assume that we're working in the two-player zero-sum setting and it does extend in some cases to other situations as well. So, our goal is to find an approximate Nash equilibrium. We're going to measure performance in terms of exploitability. You can think of it as, distance from a Nash equilibrium, it's how well we would do against a worst-case adversary relative to if we had played a Nash equilibrium instead. So, how exploitable we are Oncasinogames Canada?
I would argue that exploitability is actually extremely important and has been overlooked in the AI community as a whole. I think two recent man-machine matches actually really highlight this. One is the OpenAI, one versus one Dota2 Matches that you might have heard about, and the other is Fan Hui versus AlphaGo. In the OpenAI Matches, they made this AI that was able to beat top humans in one versus one Dota2 over three games. But after they won against the top human, they actually opened it up to the public and they invited random mediocre players to play against it to see if they could find any weaknesses. In fact, pretty quickly within a few thousand games, weak players were able to find certain tricks that they could basically fool the AI and figured out how to exploit it and beat it. Also, in Fan Hui versus AlphaGo, so they famously beat Fan Hui 5-0. But then after they published the Nature paper, they invited him to play several more matches against it to see if he could find out any weaknesses in the AI. In fact, he was able to find weaknesses where he was able to consistently beat the AI and they had to patch this before they released on the Nature. So, I think what this really demonstrates is that it's not enough to beat top humans in three or five or even 10 games. You really have to be able to consistently beat top humans, especially if you want to deploy an AI into the real world. If you're Microsoft and you're trying to deploy this products with real users, there's millions or billions of them, if there's a weakness, they're going to find it. But with the [inaudible] , we played the top humans not just in three or five hands of poker, we played them in 120,000 hands of poker over the course of 20 days. That whole time, all four players working as a team to try to exploit the AI in any way they could find. In fact, actually, I had lunch with one of the players, just a couple months ago. He said that the thing they found most shocking about the competition is that, at the end of each day, we gave them a log of all the hands that were played and we told them what the bot had on each hand that was played. This is big because in poker, a big part of the game is actually keeping your strategy hidden. If you fold, your opponents does see what cards you have. In fact, even if you don't fold but you lose the hand, you still don't show your cards are. So, you only see your opponent's hand about 20 percent, 25 percent of the time. So, like poker players will sometimes even call, just to see what their opponent had. But here, we're just giving them that information. We're telling them what the bot had on every single hand that it played. So, they didn't have to worry about that part all, and they found it absolutely amazing that they could not figure out how to exploit the AI, even though we were showing them the hands that the bot was playing every single time and the bot strategy wasn't really changing that much between days. All right, so I think exploitability is extremely important. I think has been overlooked by the AI community, and this is telling that the imperfect information game solving community has focused on throughout its existence. All right, so now, I want to get to the example of why imperfect information games are hard. I'm going to talk about a simple example that I call a Coin Toss. It starts with the coin flip. So, the coin is flipped that lands heads or tails with 50-50 probability. Player one is going to observe the outcome of the coin toss. Player two is not. So, after this coin lands, Player one has a choice. They can either sell the coin or they can choose play. We'll say, if they choose sell this to some separate subgame, the details of which are not important. The only thing that's important is the expected value. So, we'll say, if the coin landed heads, then the coin was lucky and they can sell it for 0.50 cents. On the other hand, if the coin landed it tails, we'll say it's unlucky and Player one loses 0.50 cents by selling it. On the other hand, they could choose play, and if they choose play, then it leads to Player two, and Player two has to then guess how the coin landed without having observed how it actually landed. So, if they guess correctly that is Player two guesses heads and the coin actually landed heads, then Player one is going to lose one dollar and Player two is going to gain one dollar. Here, the payoffs are shown for Player one because this is a two-player zero-sum game. So, Player two just receives the opposite payoff. Now, on the other hand, if Player two guesses incorrectly that is they guess tails and the coin actually landed heads, then Player one gains one dollar and Player two loses one dollar. You can see there's a dotted line between the two players, two nodes this signifies that Player two is in what's called an information set. This means that Player two because they did not observe how the coin landed, they do not know which of those two states they were actually in. So, why do you imagine that you are Player two in this game and yeah, so why do you imagine that you are Player two in this game, you've just observed Player one chooses play action and so you know that you are in this imperfect information subgame. So, what should you do? Should you guess heads or should you guess tails? But one option is to just always guess heads. But if you do that, that's obviously a really bad strategy because now Player two can just sell the coin when it lands heads and get 0.50 cents, and choose play when the coin lands tails and gain a dollar. So, on average they're getting 0.75 cents. On the other hand, you could always choose tails, but that's also a really bad idea because now Player two can choose play when the coin lands heads and gain a dollar and choose sell when the coin lands in tails and lose 0.50 cents but it's better than losing a dollar. So, on average, they're still getting 0.25 cents in this game. So, it turns out that the optimal strategy is to mix. It's to guess heads with 25 percent probability and tails with 75 percent probability. If you do that, then no matter what Player one does, the best they can do is just break-even, get on average zero dollar in this game. So, this is the Nash equilibrium strategy for Player two in this game, at least for this subgame. But now, let's say we change the game a little bit. Let's say we changed the payoff for the sell action. So, now, an expectation Player one loses 0.50 cents for choosing sell when the coin lands heads, and gains 0.50 cents for choosing sell when the coin lands tails. Well, it's pretty easy to see that as Player two, your strategy in this subgame should now change as well. Now, you should be guessing heads with 75 percent probability and tails with 25 percent probability. But you can see what's happened here is that, by changing the expected value of the sell action, we have affected what the optimal strategy is in the play subgame. Even though the sell action is not part of the play subgame and in fact, it's not even on the path leading to the play subgame. So, this is something that happens in imperfect information games. It does not happen in perfect information games. In perfect information games, if you wanted to determine the optimal strategy in subgame, you only need to look at that subgame by itself. But in imperfect information games, you have to look at the game as a whole. So, you can think of like perfect information games is a special case where you don't have to worry about all this stuff. Imperfect information games are the more general case where this is a problem. So, what do we do? Well, it turns out that we don't actually have to know the strategy for the entire game as a whole. I mentioned that this sell action leads to a subgame, where both players might take actions. But you don't have to worry about that, the only that really matters for determining the optimal strategy in this play subgame, is the expected value of Player one choosing sell. So, what we can do is try to estimate what that value is to Player one, and if we have that, then we can determine the optimal strategy in the play subgame. So, that's what we actually did in [inaudible]. We also have a theorem that says, "If this estimate is within delta of the true Nash equilibrium value, then we can solve for the play subgame and get within delta of the Nash Equilibrium." So, in the [inaudible] , we actually do this. We have this massive game which is simply way too large to solve upfront. So, we come up with a really good strategy just for the early part of the game, and we estimate what the optimal strategy is and what the expected values are in the later parts of the game. Now, when we're actually playing, we find ourselves in a particular subgame, we come up with a much better strategy for that particular subgame using information about the expected values from the other subgames. Then, we repeat this process, we just come up with a really good strategy for that early parts that are coming up and just estimate how to play in the later parts. We find ourselves in early subgame, we again compute a much better strategy and that particular subgame using information about the expected values of the other subgames. That's called nested subgame solving. This was the key breakthrough that allowed us to be top humans. So, when I, yes? >> Just [inaudible]. >> Yes, that's a great question. So, actually when we do this, this is sort of a general, how we would do this in general. But in poker, we solved the first two, there's four betting rounds. So, we solve the first two, with a pre-computed strategy. Because it's like each round grows exponentially in size. So, the first two rounds are actually pretty small. We got to the end of the second betting round, that's when we applied Subgame Solving. So, we came up with a much better strategy for the remainder of the game. We abstracted the bets. So, we want to consider all the 20,000 different bet sizes, we would just consider a small fraction of them. Then each time the opponent acted, each time they've made a bet, then we would solve a new subgame for that bet size. So, we would apply this recursive solving thing every time the opponent made an action beyond the second betting round. So, when I mention this idea of what's called Safe Subgame Solving, where we use the expected values from the other subgames, people always ask about this thing called Unsafe Subgame Solving, which is the more intuitive approach to doing this. The idea here is well, why don't we just estimate what the opponent strategy is? Let's say we can sort of like we played a bunch of hands against them or we can estimate what the Nash equilibrium is for them, and we figured out, well, they should be choosing play 80 percent of the time when the coin lands heads and play 30 percent of the time when the coin lands tails. Let's say, just for example. Now, if we assume that the opponent's playing this strategy, can we then reason about the distribution of states that we might be in and then solve optimally using that distribution. It turns out that doesn't work. So, let me give you an example of what this would look like. When the coin lands either heads or tails, we reason that we're in one of these states with 50-50 probability. Now if we observe player one choose play, we would say, okay, well, in a Nash equilibrium, we would expect player one to choose play 80 percent of the time if we were in the left state, 30 percent of the time when we're in the tail state. So, we update our belief about what state we're in using Bayes rule, and now we can reason that we're in that left state with 73 percent probability, and in that right state with 27 percent probability. Now, we would just say, well, if we assume this distribution is correct, then the optimal strategy is to always choose heads. But if we've already established that's a really bad idea, because now the opponent can simply shift to selling the coin when it lands heads and choose and play with the coin lands tails. So, the problem with this approach is that we're making an assumption about how the opponent is playing. Tell us a little about your book.
PaniK is a compilation of stories that was collected over the past four years from men and women all around the United States. People who have experienced single parenting, abortion, adoption, miscarriage, stillbirth, and other issues related to pregnancy were able to share their story to help support and inspire others. An amazing unplanned thing happened; while looking for these stories, siblings, grandparents, and even the children themselves started to submit their stories. Out of the 250 stories collected, 60 works appear in this thought provoking book. What inspired you to write this book? At the age of 20 I found myself single and pregnant. I wanted to share my story with other young women and men to show them that they are not alone. I started this project with a small bulletin on myspace.com. The very next day I received my first submission and they kept coming in. I have already received stories from 35 states and 4 countries. I know there is a need for this project due to the overwhelming support I have already received. How did you come up with the title and the unique spelling? I was trying to find a single word that described that moment when a person finds out they are unexpectedly pregnant. After collaborating with some family members and friends it clicked for me that the moment I felt was that of sheer panic. I changed the “C” to a “K” for marketing purposes and it stands for my daughters name, Kairi. How did you solicit for stories? How many stories were submitted? How many did you finally include in the final anthology? I advertised everywhere to get stories. Facebook, MySpace, email forwards, Poets and Writers magazine and other parenting magazines. I collected stories from people I knew who had some of these experiences as well. Overall, I received around 250 stories from 35 states and 4 foreign countries. 60 of the stories appear in the book. What are you doing to market your book?
As of right now, Facebook is my biggest way to reach people. I donate copies to libraries around my town and write my essay for local newspapers. How have sales been? Where have you had the most success? Sales have been steady. I have had the most success with the iPad and Kindle. How are readers/reviewers reacting to your book? So far, I have only received positive feedback, all of the reviews on Barnes and Noble and Amazon have been 5 star and very positive. I have already started on my second book, and have already received submissions so I must be doing something right! What was the biggest challenge you faced writing this book and how did you overcome it? The biggest challenge was that I took on such a big project and it seemed like it was never going to end. It was easy not to lose sight of why I was doing it- I am doing this to help others. However, it was hard to stay focused. I had a lot of help from people around me…some who sat with me editing for hours on end. I was very fortunate to be surrounded by such supportive people. What are the future plans for you and this book? I am hoping to go into high schools and colleges and share my experience of single parenthood and the amazing journey I have been on with this project. I have been lucky enough to be let into these peoples’ lives. It has been an incredible experience that I would love to share with others. Any advice for other writers/indie authors out there? I would say just not to give up. When I was younger, as young as 5, I knew I wanted to publish a book someday. It is still so surreal to me when I see it on the shelf at the bookstore. I know there are days…weeks…months…where it feels like you will never see that day but if you believe in yourself and you know what you are capable of you will be holding your book in no time! Tell us a little about your book.
This book is a series of cartoons based on observations of my three cats and the oddball things they do. They keep me mightily amused, and they are excellent models as they sleep a lot. Makes ‘em easier to draw. What inspired you to write this book? The book started out as a potential cartoon strip for the Washington Post. They ultimately passed on the project, so it became a book project. Why waste all that work? My feeling is nothing you do is wasted. As long as you’ve done the creative work, it can find a home somewhere, even if it’s not exactly the place where it started out. What are you doing to market your book? I entered the Global Ebook Awards contest and was nominated for an award for best Graphic Novel. I’ve been pitching the book to reviewers on blogs and am getting it reviewed that way. I also am sending out press releases about it, and I have a fan page of it on Facebook. How have sales been? Where have you had the most success? So far, I’ve had the most success selling the book at book signings and in local bookstores. How are readers/reviewers reacting to your book? I am getting a great response to the book from readers and reviewers. What was the biggest challenge you faced writing this book and how did you overcome it? The biggest challenge was when the publishing company I was working with on the book decided to drop the project. That hurt. Then another editor at another publishing house loved it and wanted to buy it, but again, the project was dropped. I had to overcome my disappointment and decide to pursue it on my own. Finally, it was my 11-year-old daughter who encouraged me to finish it and see it through. The lesson for me is that sometimes if at least one other person believes in what you are doing, they can help you realize your dream. What are the future plans for you and this book? Right now I’m in the promotion mode. My plan is to help it find a larger audience. What is your next project? I am working on a new cartoon strip that may become another book. I also have another story about two cats that is more plot driven—it’s an adventure story, with cats. A cat-o-drama. Have you published anything else? Yes, my first book EVERYDAY CAT EXCUSES was self published in 2007 and then was picked up by Sterling Press and republished in 2008 when buyers from Barnes and Noble saw it at the 2007 Bookexpo in New York. Any advice for other writers/indie authors out there? Be tenacious in believing in your work and getting it out there. Don’t let a negative response deter you if you really believe in your work. Know you will find your audience if you keep on believing in your book or any other creative project you have. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |