Video 0-1: Welcome, teaching mode, overview, assessment, academic integity, Campuswire, Slido, exam question gallery

0:00 Module introduction

0:03 Welcome to efficient algorithms. Today we'll talk mostly about organizational things and I'll address the prerequisite version of the module and the four credit version of the module and the second will mostly be in German. So at some point I'll make a hard cut and switch back to German. Uh for the moment in the interest of our master student from abroad um let's stay with English for a little bit. Um the goal for today is to make you all aware how this this module is organized and a bit of what we'll cover and uh I mean if it is an elective module for you if you have the choice then you should be able to decide whether you like it enough or to keep it or whether you want to do something else and we'll briefly talk about the assessments uh how you get marked So, welcome. That's me. Uh, Sebastian Vid, Sebastian Wild. Um, whatever you prefer. I'll probably react to both.

1:06 Um, my office is in in hallway 5D. Uh, the safest is to always fire an email first and see if I'm in if you want to avoid disappointment. Uh otherwise, email is the preferred way if you need anything. Um we'll soon have um another person joining my group. Um but he's not here yet. Uh I did bring one of the tutors. Finn Molman is here and maybe some of you know each other as well. Anyways, and Arsenal is the second tutor. So you made up which which group is which. Does that does that matter? Do people need to know who is leading their uh session to vote? Maybe not. I don't know.

1:58 Okay. Finistering Wednesdays and uh Arsen will do the Thursday session. Uhhuh. We'll have to kind of load balance it a bit. Okay. Um I tried to put all the important information in one slide. some of which I've mentioned before. There's the module website. Here's the link uh that you've seen before. This is where I try to keep most of the information. Anything that can be public will be on that website just because I find it myself find it more convenient not to have to log in for every little bit. Uh but also I think it's uh yeah for you as well more convenient. We'll use campus wire. what I mentioned briefly there's this pin again to join in the link uh we'll use slido which maybe you've logged into we'll come to that in a second and just to get that get that information out unless you're shouldn't be a surprise for anyone uh there will be some exam a written exam at the end of the module and that's going to be your final mark um this is pretty much prescribed by our regulations that the module mark is equal to the mark of the final exam To be allowed to uh take the exam, you have to participate in the exercise classes and there will be weekly exercise sheets that you submit in groups. More on that soon. And you'll have to reach a certain number of points from those. If you don't reach those, then you're not admitted to the exam, so you cannot take it.

3:31 It's pretty standard in uh our overall program. Um but important to mention all right note on languages. Yeah the module itself has to be in German because it is part it is a required module on our bachelor's course on data science and that one is in is in German. So uh irrespective of what I would maybe think is useful or prefer or doesn't matter uh it'll be in German and the same is true for the examinations although we'll accept English answers if you write them in English that's fine for us it's a different story for the prerequisite module I'll comment on that in a second um what we can also provide for if if there's students that want to take the module and don't take it as a prerequisite module module that want to take it for credit. For example, international students on the masters in computer science, our group that might choose this one. We can provide the exam questions as an informal translation in English. That's fine. And you can submit your answers in English. The official version has to be in German.

3:42 Language and communication

4:47 Okay. I also decided to keep the slides in English because I think it's useful. The terminology is mostly in English. Anyways, you'll see me struggle translating things once in a while. Uh, this whirlwind here is places I've spent uh my professional life and so there's also some personal confusion with some terms that will shine through at some point. Uh, you'll see. Um, besides that, translating is way easier than it used to be and so hopefully that's not a big hurdle for most people. also because the uh lectures are recorded and you can watch the recordings again with YouTube's automatic subtitles. I'm not uh uh generating those manually, but they're pretty decent. And so maybe at least slowing things down or so can help some students uh if language is an issue.

5:45 Conditional admission details

5:47 Okay, I think enough of that. Um now let me briefly speak specifically to students that have this uh conditional admission prerequisite module. So for context for everyone we have an international program masters data science that is a consecutive masters program. So you have to have a specific undergraduate module to be admitted to it. And um at least for some iterations they were students were admitted sub subject to doing this conditional admission requirement after they started. And um that means there was a gap in their undergraduate program with respect to what our masters assumes. And to fill that gap, to pluck that gap, um this module or the the conditional admission version of that module, formally they're kind of separate, um was uh was given to students to kind of fill that gap and you have to pass that exam kind of retrospectively to be admitted to your program.

6:50 So the normal version is is separate from that. The conditional uh admission version is technically separate from the lectures as well. There's a decent overlap in terms of the material that we cover, but it's not identical. And to be specific, the material that counts for the prerequisite students is linked here on this separate page. and it's exactly what is listed there that is relevant for the exam for for the prerequisite students. Uh there's a um a YouTube playlist of videos from old recordings from my previous employment in Liverpool where they were in English. And so these cover most of the material from the lectures and then there's a few selected book chapters from this um a standard algorithm book that covers the rest that wasn't covered in any of the videos that I had. Now uh what you don't need to do is whatever I said about these exercise sheets in admission to the exam that does not apply to the prerequisite module there. you can just show up for the exam and anything you do in preparation before that is is uh uh under your control. Uh what we'll try to do this year is offer a bit more of support for exercise problems at least have the sheets prepared in English.

7:16 Exam material and resources

7:27 Curriculum topics

8:12 We'll we'll try that and see uh how the uh demand for that is and uh we'll also have some old exams from last year. Now it's the second time I'm doing the module. Last year a lot of things were um without prior versions a bit harder to prepare for the exam. Um but this year we'll release on on Elas the old lecture questions the old exam questions and then prerequisite students also see a bit what the exam is ahead of time. Okay. But again important bit is the material is exactly what's listed on this on this subpage of the module page. It's what's in the videos and what's in those lecture and in these book chapters and nothing else. Okay.

8:46 Slido and lecture pedagogy

9:09 Uh this came up last year. I don't know if this is still relevant. You cannot do both modules. Um if if you're in the situation that you're given this as a conditional admission prerequisite, then it means you're you're deemed to have taken this in undergrad and then you can't get credit for the same module again. Uh this this is similar as if some of you were doing your bachelor's in my book and stay for the masters, you can't get the same module uh as credits in both programs.

9:43 Um yeah, this this point is important to stress. Uh unfortunately because it's a a conditional admission module, we don't get any extra support for the university. This is kind of not official teaching. We're not getting extra credit for doing extra work and we don't get in particularly particular we don't get extra funding for more tutors. So we can't really support a second uh exercise class with marked submissions for the exam for the for the exercise problems or other exam prep. What we'll try is to see how the load balance works out and if our tutors have some some capacity left then they'll offer to look over submissions from prerequisite students as well. But this is something I can't do a firm promise on. But approach us um and we'll we'll come back to how uh what other methods you have to involve to involve yourself with the module. You have to be a bit proactive on on this one.

10:42 And in particular, there's a a question and answer forum uh that you you're you're well welcome welcome to use. I mean you're the modules are formally separate the two versions of this uh of this class but you're very h welcome and I'm happy if you mingle with the students and kind of work together towards learning the stuff that this what this is about.

11:15 Let me make some like a a little a little do detour and like for for the few international students that are in the audience but maybe some more are watching as well in the recording. I think we're uh we're all very happy that you're here and the fact that we can't support you through this module is is not something I like. Uh we would love to offer offer better support to get you up to speed for the exams on this. Uh it's tough with the resources that we have, but we'll we'll try and uh you're welcome. Um I hope you settle in our in my home country, in our home country. If if anything, this country needs good people to come to us. It needs immigration. There's a lot of uh of political push against that. But I I want to take a firm stance on the opposite side. We need immigration. We need good people to come here. And you're you're welcome. Um, but we're bound to what the resources allow us to do for the module. And so we'll we'll have to work our our way through this.

12:24 Let's talk a bit about Slido. Um, the reason I'm using the system is for a simple psychological trick. You sit there for 90 minutes in a standard lecture in TV mode. And what that does to your brain is that you become a bit passive. And Slido is a very simple hack, uh, a psychological hack to get you, uh, back awake. And that's pretty much mostly what I'll use it for. There's more pedagogy behind it. But I think this is the biggest influence really. It's a quick way to check uh, where people are and so on. It's it's an easy way for you to ask questions. You can ask questions anonymously if you prefer. Uh all of this is great. Um there's usually a few students that are happy to ask questions just uh or verbally, but then many more that have smart thoughts and you're welcome to use slider to to utter them to the the rest of the class.

13:17 So there's two main features that we'll use. One is the uh the polls where I'll ask you questions, usually multiple choice, and on your devices you can click and answer. And the other is that's the that's the push version. I'm pushing a question to you. And the other is pull. You can send questions. I mean, you can try it if you want. Um that's that's the other option. You can just uh use that any time. I might not I might try to see what happens, but might decide to wait for a good moment in the material to take a pause and answer questions.

13:55 Audience response system

13:58 And so once in a while on the slides this will pop up which is the point where um I'll start the interaction on slido and you can vote on the thing and the way it looks in slido is this. So you see in this case you immediately see what the others vote. I'll also keep an eye on this number up in the bottom that tells me how many people are well paying attention physically in the room is not the same. Um, now minus the few that might have uh not brought some device that can vote.

14:41 Okay, pretty half half. That's uh okay. Maybe now few more people have used these things before. This is a very neutral question. It doesn't ask about whether you liked it or not or whether it was useful.

15:12 Okay. Something like this. Let's call it 33. Uh so it it seems to work. and is pretty intuitive. Now, that also ties into um my approach to lectures. I've already commented on or I told you what I'm what I'm planning to do in terms of recordings, etc. I've experimented over the years with different formats. There's all these cool uh lip teaching and that that that was the new thing for many many years. And I tried it a bunch and it never really worked for me. Um maybe others are are better with this. So for me the main thing is a a synchronous life lecture because for the simple reason that it keeps you focused on some specific topic at least once or twice a week if nothing else. Um, if you just have recordings or you just have book chapters or whatever else interactive formats, it is incredibly hard to stay on topic and stay with the teaching for a longer period of time. I've tried that as kind of on the other side of of watching MOO's massive open online courses even not sure if that's still a thing that term. Um, that was already before the pandemic a hot topic and now the pandemic everyone was teaching online. fine. Uh but I I noticed the only thing that really works for me in in my teaching is doing things synchronously so that people have a minimum engagement time regularly.

15:16 Teaching philosophy

16:44 I think this is overall still the best but you get the recordings in case it doesn't work at all for you. You you decide. I want to treat you as the adults that you are. And if you think you have a better model that works for you, that's fine by fine by me.

17:00 I've talked about uh asking questions anonymously. That's what Slido is for. I think this is really useful and important. Um I want to hear from as many of you as possible and that's the easiest option for doing that is via some online tool. I'll come to the question in a sec. Okay. And then keeping you awake a little bit. So what does that mean for the module apart from what I've already mentioned? Uh we'll have the lectures. That's where you're currently sitting in uh with the questions and we'll have the tutorials. Then we'll use campus wire as a everything that happens outside of class for any asynchronous communication there. I mean, you're always welcome uh to come come and see me if you have questions, but I'll often redirect you to campus wire for things if they also concern other students. I I'll tell you more what this looks like, but it's a bit like Reddit. So, you can ask questions and answer them. And the thing is that if you ask there, everyone gets to see the answer and everyone profits.

17:22 Module components overview

18:06 Uh, a little more experimental thing is an exam question gallery. So, I'll I'll try to crowdsource making good exam questions to all of you. And um the deal is that if you do a good job there, then maybe one of these questions pops up on the actual exam and you know it up front. which is uh as a useful thing. And then of course the final exam um by other regulations is the only thing that gives us uh the final mark.

18:40 Slido demonstration

18:41 Okay, just for proof of concept, let's go to slido. Someone sent this um I delivered this question. It seems to work. So I would take this off and let me go back in case of a real question might spend some more time. Okay, let's also talk a little bit about the content of the module at this point. Um we'll we'll basically look at the basic algorithms and data structures that many other things build on. And while we focus on on the practical methods meaning the methods that have efficient running times or have small running times. uh the focus is is much on the methods and techniques so that you're able to um also generalize them, use them for other problems that might pop up wherever you end up working. Uh that uh is often something that people report back even if they're in some industry job that doesn't seem to work much with algorithms. Um the algorithms lectures was often something that they looked back at and found it useful.

18:51 Content overview

18:52 Final assessment details

18:57 Tutorial and exercise policy

19:04 AI usage and ethics

20:01 um if nothing else than to get through the tech interviews. All right. Um what this looks like uh in terms of individual sections. So this is is my way of breaking up the material in roughly weeksiz chunks. So each is roughly two lectures. It's going to be true on average, not exactly. Uh we're currently here and then we'll look at these things. Uh more more on that when we get to it.

20:39 Let's not skip the assessment. Uh you'll have to have some final mark. So everyone on the prerequisite module, this is this is the German version of the exam. uh it won't concern you un uh except for the dates. These are preliminary um likely needs to be stressed a lot more. This might still move uh but in case you want to keep keep the dates free, these are likely the dates of the final exam.

22:22 relative standard in my exper. this is relevant for everyone. So, let's switch back to English. Actually, it's a it's a cool and exciting time. Uh, I'm using chat assistants for various things. They're pretty cool. They're good for some things. They're stupid for other things. There's one one thing where they're really harmful, particularly because they work reasonably well, and that's for solving exercise problems. Uh especially if they're fairly narrow and maybe programming focused. Uh the usual LLM based tools work really well. Um and maybe just because they've seen a similar solution on the web, it doesn't really matter whether it's impressive AI achievements or not. uh what it does is that it kills the effect for you. If you take our exercise problems which is the uh admission to the exam it's not what you're marked on. I mean needs needless to say during the exam you'll be on your own and you won't have any assistance from any tools. Uh so it's just your brain and your and a pen and a piece of paper. That's what we want to assess because we still we still I'm I'm deeply convinced that we need skilled people for various reasons. Uh and uh not the least is that AIS make horrendous mistakes sometimes and we want some I don't know uh most most sane humans would not do certain things and so I I want to rely on our society being built on mostly sane human beings also in the future but that means you need to acquire certain skills if you zoom out a bit I guess I hope that's why you're here in university anyways. You get some fancy piece of paper at the end that might open some doors. Uh but really what you should want to take out of all of this is that you you're society approved taking a few years out of your uh usual life of working to work on fun things and learn new things uh learn some fun things. So don't take that fun away by um solving the exercises with too much AI help.

26:46 It can be uh it can be useful to learn how to use AI tools, but don't use it for the exercise problems. It's working against you in the end.

27:00 You're still encouraged to use them. Um for example, for a replacement for learning about basic facts about some topic, that's totally fine. It's also fine to use them as a uh a spell checker on steroids. uh what's not okay is to use them to generate large parts of your assi assessment or your assignments and in particular not mention that you've used them. So if you use them for something at least give them credit and list what you've done uh but don't make them co-authors. Okay, that's what uh the major conferences in computer science settled on as the rules for AI in research.

27:46 I know this is all pretty um reasonable in the abstract and I guess is not too surprising that this is something I would like like you not to do or like you to do. Um but here's another uh thing that plays into this if you had to guess. So this is predictor here means um the question says what do you think is the number one predictor of whether a student cheats in assessments and predictor means you're you're collecting data from many people and then you're trying to find correlations and it's correlation strictly not causation we can't from the data I know uh we can't tell uh but if you were to ask people various things um and I know I'm using this question. Um, some people have seen it before. I'm using this in all my lectures. Uh, what is the number one thing about a person that correlates best with whether or not they cheat on assessments? Self-reported cheating.

27:56 Cheating predictors

29:01 Procrastination and management

29:10 I'll show you what the others send in in a second, but I don't want to uh stop your thinking straight away. Well done. Okay, let me show you uh what we got. Lots of different answers.

30:08 Discussion on student feedback

30:19 Grandma, is that grandma or is it grammar? Uh, okay. Whatever. Uh, both are interesting guesses. I think uh not in lectures. Lack of time, time pressure. Same solution. Not going to lectures, hence exercise is too hard. Yeah. Assignment being too good.

30:44 I'll take that. That's okay. Uh when it's easier to do formality. Okay. Again, chromatic. H. Okay. I'll ponder over this. Uh the right one's on the slide. Um, and um, if you want to hear the full story, there's a YouTube video link that you can watch. The number one predictor is procrastination. So whether or not people have trouble managing their time and constantly end up doing things last minute plus spending time on things, the procrastination part is spending time on things that they afterwards regret and would have rather preferred they had spent otherwise. And so this I mean it's probably not a causation in a strict sense but what it often what I take from this is that uh if you're getting desperate because you need to submit something right now otherwise you'll all these bad things happen then you start cheating because you run out of better options. Now, if you're already in this desperate mode, there's very little I can really do. The thing that you should do is not get into that mode whenever you can. And the number one thing you can do for that is good time management. This is boring, and I don't want to get into a management seminar here. Uh, but maybe you should think a little bit about whether you can plan your time um a bit more consciously. There's good tools for that. The further you go in your professional lives, the harder these things will get.

31:01 Campuswire tool guide

32:23 Um, so it's a good skill to have and there's very very easy, not at all sophisticated tools that you can use. So that's that's a a fun thing to keep in mind. Um, of course we've all been there. Um, within this year, there were two paper deadlines where I've written papers late at night. As this stuff happens, sometimes you can't avoid it, but uh, you should try to not always have it happen.

32:54 Collaborative learning guidelines

32:57 Okay. Uh for the rest of this module, I want to give uh a bit of an overview of the different tools um that I'm I'm using. I'll try to go a little quicker because maybe these are all intuitive enough. Uh number one is Campus Wire. That's this question and answer forum, the Reddit for lectures. If you want, um you can join with that link and the pin. You can do it now if you want. Uh there's two features that we'll use. Um one is called class feed and that's the Q&A part. Um there's also chat rooms that you can use if you want. Um I'm not going to use them a lot. They're essentially like a a typical chat room with the chaos that comes with that. Um, the class feed is what I'll mostly use and you can it's a combination of sending announcements from us to you and you can just format them in markdown and use latte formulas and these things.

33:51 It's a bit more nerd compatible compared to alas or the other tools that we have. Um, and you can also ask questions from your site. And again, you can do this anonymously or with your name. So specifically this uh the class feed is a collaborative Q&A. I want you all to also answer other people's questions and you can ask a question in public or in private. The default should be to ask in public unless you think it kind of spoils some assessment or is is something that's really just relevant for yourself and uh you can ask a private question there. That's fine. But this is essentially like sending me a private email except that also the tutors can answer. So it's often faster. The public question should be pretty much anything else. In particular, anything that is a question about the material of the lecture or some question you would ask me after a class or ask in the tutorials often is is better suited to ask on campus wire basically for two reasons. Um and that's uh the goals that we have for this. Everyone gets the same answer.

35:01 That's fairer. If you ask me in some private office hour and I give you one answer and then the next person asks me the same thing a week later and I give a slightly different answer, this is not so great. I prefer you to have the same information. And then the other reason is you can also answer questions which is where this really becomes interesting because uh if you think about a question and make an effort to formulate an answer and you can do this whatever language okay it doesn't matter you can also take uh little uh photos of some handwritten handwritten stuff uh this all rel relatively flexible they have a mobile app which is not perfect but it's usable If you think about other people's questions and answer them that can that's going to help you a lot more to learn things is an old saying among teachers that you only really understand something once you try to teach it to someone else and it's just it's so true.

36:00 It's boring as a statement but it's so true. So um use it as as that. Uh the promise from our site will be that's um on how we will so all the tutors uh will also be on there to answer questions but also other students. So what we'll promise to do is if the if some student answers a question we'll read through it and if it's if it's correct I'll we'll mark it with some endorsement. There's a a feature of the system to say this answer has been checked by an instructor. So even if it's a student answer, you can rely on it. Uh or if something's missing, we'll just write a little addition to it. Okay.

36:45 We'll also to encourage this, wait a bit before answering any question in case someone wants to chip in. Uh and I'll I'll usually ruthlessly redirect you to Campus Wire. If you send me a a private email with some question that could have been a public question, I'll just send back Campus Wire. Okay? So you can save us this email back and forth uh if you directly go there. Uh for anything that's private or personal that only concerns you, happy to get your emails.

37:18 Yeah, also upvote answers and questions. You'll find this is easy to use. There's still IAS around and Marvin. So, ELAS will use for anything that is official or where the where the data is is sensitive. So, any uh marks or your submissions of uh tutorial exercise sheets, those will be on alas and you'll get your your results there from the marking of the submissions and we'll also eventually put the exam marks there. Uh for anything else, I'll try to use the more convenient systems. Uh where data is sensitive, we'll stick to ELS. Okay.

37:21 System usage summaries

38:04 Exam question gallery

38:11 Okay. Almost the last bit. This uh ominous exam question gallery. Uh maybe it's a bit early in the term to really discuss this in detail. Uh again the goal is if you think about how to assess a certain topic that's actually very good preparation for answering questions that I come up with or we come up with. So thinking about exam questions is another way to get you actively engaged with the material and thinking about good exam question is actually really hard. So, um I think it's a fun process to go through and it also forces you to think about what could be on the exam which at some point later in the term you'll do anyways. But my my uh point here is do this while we go along while you're sitting here. Anyways, think about how would you assess this topic and write down the question already during the term.

39:08 And again this is open to all the students on either of the module versions. Um and it's some version of of a latte online system like Overleaf but one that does allow an unbounded number of collaborators. So you'll learn a bit of latte on the way or you ignore it and uh someone else fixes the type setting for you. So um all I can really say is try try this. I think it's useful for you and the promise is if you come up with good questions um we'll select some to put them on the actual exam. I might not tell you which ones up front but uh you increase your chances and knowing a pool of questions that will be selected from I think is is an advantage for the exam.

39:55 Scientific foundations

40:01 Okay, the last bit uh it's just a little in terms of the topic setting uh the focus on the right thing. Uh first of all, it's a university program. So the focus is on foundational scientific skills, understandings, uh connections. It's not just a vocational training course. It's not just about teaching you how to have a the skills on your next job or teaching you the current JavaScript framework that gets you that web developer pro job. Uh it's about the foundations that last over several decades. And for this module specifically, it's part of theoretical computer science. Uh the hope is that we'll get a bit more towards this part. Let me just read out this one. There's an XKCD comic. In case you don't know these, there'll potentially be more of those in the lecture slides. They're pretty nerdy, but uh good ones. So, the world is so complicated. The more I learn, the less clear every anything gets. There are too many ideas and arguments to pick and choose from.

41:08 How can I trust myself to know the truth about anything? And if everything I know is so shaky, what on earth am I doing teaching? Ah, I guess you just do your best. No one can impart perfect universal truths to their students except math teachers. Thank you. So we'll focus on the universal truths of practical algorithms which sometimes means making a slightly more abstract statement in an abstract model of computation and all these things as opposed to just saying this currently runs fine on my machine. I don't bother. Why? And then next year it doesn't and I'm down here or well the other way around. Uh what this means is that we'll need a bit of uh math background and that's what we'll cover in the next unit.