Decision trees have long been a prevalent area within machine learning. With streaming data environments as well as large datasets becoming increasingly common, researchers have developed decision tree algorithms adapted to streaming data. One such algorithm is SPDT, which approaches the streaming data problem by making use of workers on a network combined with a dynamic histogram approximation of the data. There exist several implementations for decision trees on GPU, but those are uncommon in a streaming data setting. In this research, conducted at RISE SICS, the possibilities of accelerating the SPDT algorithm on GPU is investigated. An implementation is successfully created using the CUDA platform. The implementation uses a set number of data samples per layer to better fit the GPU platform. Experiments were conducted to investigate the impact on both accuracy and speed. It is found that the GPU implementation performs as well as the CPU implementation in terms of accuracy, suggesting that using small subsets of the data in each layer is sufficient for making accurate split decisions. The GPU implementation is found to be up to 113 times faster than the reference Scala CPU implementation for one of the tested datasets, and 13 times faster on average over all the tested datasets. Weak parts of the implementation are identified, and further improvements are suggested to increase both accuracy and runtime performance. / Beslutsträd har länge varit ett betydande område inom maskininlärning. Strömmandedata och stora dataset har blivit allt vanligare, vilket har lett till att forskare utvecklat algoritmer för beslutsträd anpassade till dessa miljöer. En av dessa algoritmer är SPDT. Denna algoritm använder sig av flera arbetare i ett nätverk kombinerat med en dynamisk histogram-representation av data. Det existerar flera implementationer av beslutsträd på grafikkort, men inte många för strömmande data. I detta forskningsarbete, utfört på RISE SICS, undersöks möjligheten att snabba upp SPDT genom att accelerera beräkningar med hjälp av grafikkort. En lyckad implementation skriven i CUDA beskrivs. Implementationen anpassar sig till grafikkortsplattformen genom att använda sig utav ett bestämt antal datapunkter per lager. Experiment som undersöker effekten på noggrannhet och hastighet har genomförts. Resultaten visar att GPU-implementationen presterar lika väl som CPU-implementationen vad gäller noggrannhet, vilket påvisar att användandet av en mindre del av data i varje lager är tillräckligt för goda resultat. GPU-implementationen är upp till 113 gånger snabbare jämfört med en existerande CPU-implementation skriven i Scala, och är i medel 13 gånger snabbare. Svagheter i implementationen identifieras, och vidare förbättringar till implementationen föreslås för att förbättra både noggrannhet och hastighetsprestanda.

In this work, a model of computation for shared memory parallelism is presented. To address fundamental constraints of modern memory systems, the presented model constrains how parallelism interacts with memory access patterns and in doing so provides a method for design and analysis of algorithms that estimates reliable execution time based on a few architectural parameters. This model is presented as an alternative to modern thread based models that focus on computational concurrency but rely on reactive hardware policies to hide and amortize memory latency. Since modern processors use reactive mechanisms and heuristics to deduce the data access requirement of computations, the memory access costs of these threaded programs may be difficult to predict reliably. This research presents the Queue Streaming Model (QSM) that aims to address these shortcomings by providing a prescriptive mechanism to achieve latency-amortized and predictable-cost data access. Further, the work presents application of the QSM to algorithms commonly used in a number of applications. These algorithms include structured regular computations represented by merge sort, unstructured irregular computations represented by sparse matrix dense vector multiplication, and dynamic computations represented by MapReduce. The analysis of these algorithms reveal architectural tradeoffs between memory system bottlenecks and algorithm design. The techniques described in this dissertation reveal a general software approach that could be used to construct more general irregular applications, provided they can be transformed into a relational query form. It demonstrates that the QSM can be used to design algorithms that enhance utilization of memory system resources by structuring concurrency and memory accesses such that system bandwidths are balanced and latency is amortized. Finally, the benefit of applying the QSM algorithm to the Euler inviscid flow solver is demonstrated through experiments on the Intel(R) Xeon(R) E5-2680 v2 processor using ten cores. The transformation produced a speed-up of 25% over an optimized OpenMP implementation having identical computational structure.

[pt] A presente dissertação propõe uma discussão a respeito das transformações na produção e no consumo audiovisual, a partir do avanço das tecnologias digitais. Para isso, se refletirá sobre as múltiplas formas de consumir produtos audiovisuais, apresentando um breve quadro panorâmico de transformações na produção e veiculação dos produtos audiovisuais, desde os meios tradicionais de broadcast até o streaming. Uma das questões da pesquisa é compreender como a introdução de novas ferramentas de produção, novos dispositivos, novos suportes e novas plataformas de distribuição, configuram novas rotinas e novos modos de ver ao permitir aos espectadores um amplo espectro de escolhas dos modos de assistir (lugares e telas) e diferentes configurações de consumo, sejam individuais e/ou coletivas, presenciais e/ou virtuais. Ao mesmo tempo, dependendo da ocasião e da geração, encontram-se novas formas convivendo com antigas práticas de assistir audiovisual. / [en] The present dissertation proposes a discussion about the transformation in the production and the audiovisual consumption, from the advance of the digital technologies. To do so, it will reflect on the multiple ways of consuming audiovisual products, presenting a panoramic board of the transformations in the production and delivery of audiovisual products, from traditional broadcast media to streaming. One of the research questions is to understand how the introduction of new production tools, new devices, new media and new distribution platforms, configure new routines and new ways of viewing by allowing viewers a wide range of ways to watch (places and screens) and different configurations of consumption, whether individual and / or collective, face-to-face and / or virtual. At the same time, depending on the occasion and the generation, there are new ways of living with old audiovisual practices.

no / Peer-to-Peer (P2P) has become a popular live streaming delivery technology owing to its scalability and low cost. P2P streaming systems often employ multi-channels to deliver streaming to users simultaneously, which leads to a great challenge of allocating server resources among these channels appropriately. Most existing P2P systems resort to over-allocating server resources to different channels, which results in low-efficiency and high-cost. To allocate server resources to different channels efficiently, we propose a dynamic resource allocation algorithm based on a streaming quality model for P2P live streaming systems. This algorithm can improve the channel streaming quality for multi-channel P2P live streaming system and also guarantees the streaming quality of the channels under extreme Internet conditions. In an experiment, the proposed algorithm is validated by the trace data.

"The Internet has assumed the role of the underlying communication network for applications such as file transfer, electronic mail, Web browsing and multimedia streaming. Multimedia streaming, in particular, is growing with the growth in power and connectivity of today's computers. These Internet applications have a variety of network service requirements and traffic characteristics, which presents new challenges to the single best-effort service of today's Internet. TCP, the de facto Internet transport protocol, has been successful in satisfying the needs of traditional Internet applications, but fails to satisfy the increasingly popular delay sensitive multimedia applications. Streaming applications often use UDP without a proper congestion avoidance mechanisms, threatening the well-being of the Internet. This dissertation presents an IP router traffic management mechanism, referred to as Crimson, that can be seamlessly deployed in the current Internet to protect well-behaving traffic from misbehaving traffic and support Quality of Service (QoS) requirements of delay sensitive multimedia applications as well as traditional Internet applications. In addition, as a means to enhance Internet support for multimedia streaming, this dissertation report presents design and evaluation of a TCP-Friendly and streaming-friendly transport protocol called the Multimedia Transport Protocol (MTP). Through a simulation study this report shows the Crimson network efficiently handles network congestion and minimizes queuing delay while providing affordable fairness protection from misbehaving flows over a wide range of traffic conditions. In addition, our results show that MTP offers streaming performance comparable to that provided by UDP, while doing so under a TCP-Friendly rate."

The record industry has been around for just over 100 years and it’s an industry that mostly has been growing and growing ever since its inception. New technology has always pushed the industry forwards and helped it grow bigger and bigger with almost each year. But when the internet came in the early 2000’s it changed everything. It caused a stir in the industry when the music consumers now had the possibility to rip CD's, download music files illegally and share them with the world for free. In desperate actions to try to save the record labels from going bankrupt they did all in their power to fight the change. What eventually happened was that other actors came in and saved the business with services that embraced the change from a physical way of consuming music, to a new digital one.   This is a study of how that change has been perceived and affected the record labels and their ways of working. By using Paul Hirschs modell charting the organization of the pop music industry and the four processes that Tschmuck describes as the core activities of a record label, we’ve interviewed six record labels in the Stockholm area in an attempt to describe the new core activities and organization of the industry now that the digitization has changed the way the customers consume music.   Our findings show that the value added chain that Hirsch modelled has changed drastically, with several actors losing importance in it, while others actors has traded places. The core activities that Tschmuck described remains the same, but the ways they are executed has become more effective thanks to the new technology.

En undersökning av hur online-exponering påverkar kreativa processer. Genom att arbeta i en öppen live-stream har Jerker Österlind inför publik skapat en EP under 30 Twitch-sändningar. Denna text utforskar effekterna av att bjuda in åskådare till vad som brukar ske bakom stängda dörrar, hur det musikaliska resultatet påverkas, hur arbetsprocessen förändras och funderingar om hur livestreaming kan ge nya möjligheter för marknadsföring och karriär. https://open.spotify.com/album/18uGeyxD00Ssayi1exCCJo?si=qbFel0xUTaeF4rXiwlR2GQ / An inquiry about the impacts of online exposure on creative processes. By working in a public livestream, Jerker Österlind has made an EP in front of an audience during 30 Twitch broadcasts. This text explores the effects of inviting an audience to things that usually take place behind closed doors, how the musical result is affected, how work methods are changed and how livestreaming can open up new opportunities for marketing and careers.  https://open.spotify.com/album/18uGeyxD00Ssayi1exCCJo?si=qbFel0xUTaeF4rXiwlR2GQ / <p>Arbetet presenteras i form av en EP med fem spår. Alla instrument är spelade av J. Österlind.</p><p>Musiken kan höras här:</p><p>https://open.spotify.com/album/18uGeyxD00Ssayi1exCCJo?si=qbFel0xUTaeF4rXiwlR2GQ</p><p>och här:</p><p>https://mudbody.bandcamp.com/album/underachieve-ep</p>

Konsumtionen av video på Internet har ökat det senaste decenniet och trenden är förväntad att fortsätta stiga. Strömmande video är en ofta använd lösning för att se video på Internet. Tekniken tillåter användaren att se film medan den laddas ned. Adaptive streaming är en teknik som låter videospelaren ändra bit-raten på videon baserat på användarens tillgängliga bandbredd. En annan lösning för att göra användarens upplevelse mer skräddarsydd är ickelinjära videor. Den här typen av videor kan spelas ur sekvens. Som exempel kan en tittare välja slutet i en film eller välja vägen genom en utställning. Examensarbetet presenterar designen och implementeringen av en ny struktur för ickelinjär video. Denna struktur används av videospelaren för nedladdning och visning av adaptiv video avsedd för ickelinjära filmer, som sparas på en server. Mediaskapare kommer också att ha lättare att både visualisera och skapa filmer på ett enhetligt sätt. Det här examensarbetet presenterar i detalj de modifikationer på Adobes Open Source Media Framework och Strobe Media Playback som tillåter spelning av ickelinjära videor. Ändringarna innefattar ett användargränssnitt för ickelinjär media. Back-end implementeringarna innefattar bufferthantering för parallell nedladdning och tekniker för att hantera den nya strukturen. Slutligen presenterar det här examensarbetet proof-of-concept-tester för att styrka funktionaliteten av designen och implementeringen. Rapporten avslutas med en diskussion om framtida arbete inom fältet. / Video consumption on the Internet has been growing over the last decade and is expected to continue to increase. Video streaming is a widely used approach to viewing video on the Internet, which allows users to watch the video while it is being downloaded. Adaptive streaming is a video streaming technique that allows the player to change the downloading video’s bit rate depending on the user’s available bandwidth. Another approach to a more personal viewing experience is non-linear videos. These videos can be played in a non-sequential order. For example, a viewer can be given the choice of what ending to watch in a movie, or the path through an exhibition. This thesis will present the design and implementation of a novel structure for non-linear video. This structure is used by a video player for downloading and viewing an adaptive video intended for non-linear viewing, stored at some server. Media creators will also have an easier time to both visualize and create uniform video experiences. This thesis presents modifications to Adobe’s Open Source Media Framework and Strobe Media Playback which allow playing non-linear video. Presented in this thesis are the design and implementation details. Changes in the player include a user interface for non-linear media. The changes to the back-end include buffer management for parallel downloading and techniques to handle the new structure. Finally, this thesis presents proof-of-concept validation tests that show the functionality of the design and implementation. The thesis is concluded with a discussion of future work in this area.

Den stora tillgängligheten till Internet har inneburit att allt fler människor börjat kommunicera och interagera online, både via sociala medier och onlinespel. Antalet nya plattformar där människor med samma intresse kan mötas och interagera ökar hela tiden. En av alla dessa plattformar är Twitch där individer kan välja att sända live när de spelar dator- och tv-spel medan andra kan välja att agera publik. Utifrån grundad teori undersöker vi hur personer upplever att streama via Twitch och vilka de mest centrala aspekterna av detta är i syfte att öka förståelsen kring detta relativt nya fenomen. Genom intervjuer, frågeformulär och observationer lyfter studien fram hur dessa personer upplever bland annat gemenskap, krav, frustration och näthat genom streamingen. Resultatet visar bland annat på att det är upplevelsen av gemenskap som är det mest centrala inom fenomenet och i de flesta fall det som gör att personerna väljer att ägna sig åt streaming. Utifrån Giddens teori om det senmoderna samhället kopplas fenomenet till ett nutida sammanhang och med Goffmans teorier om roller belyses samspelet mellan streamarna och deras tittare.

<p>The ever increasing demand for video content directed the focus of researchfrom traditional server-based schemes to peer-to-peer systems for videodelivery. In such systems, video data is delivered to the users by utilizing theresources of the users themselves, leading to a potentially scalable solution.Users connect to each other, forming a p2p overlay network on top of theInternet and exchange the video segments among themselves. The performanceof a p2p system is characterized by its capability to deliver the videocontent to all peers without errors and with the smallest possible delay. Thisconstitutes a challenge since peers dynamically join and leave the overlay andalso contribute different amounts of resources to the system.The contribution of this thesis lies in two areas. The first area is theperformance evaluation of the most prominent p2p streaming architectures.We study the streaming quality in multiple-tree-based systems. We derivemodels to evaluate the stability of a multiple tree overlay in dynamic scenariosand the efficiency of the data distribution over the multiple trees. Then, westudy the data propagation in mesh-based overlays. We develop a generalframework for the evaluation of forwarding algorithms in such overlays anduse this framework to evaluate the performance of four different algorithms.The second area of the thesis is a study of streaming in heterogeneous p2poverlays. The streaming quality depends on the aggregate resources that peerscontribute to the system: low average contribution leads to low streamingquality. Therefore, maintaining high streaming quality requires mechanismsthat either prohibit non-contributing peers or encourage contribution. In thisthesis we investigate both approaches. For the former, we derive a model tocapture the evolution of available capacity in an overlay and propose simpleadmission control mechanisms to avoid capacity drainage. For the latter, inour last work, we propose a novel incentive mechanism that maximizes thestreaming quality in an overlay by encouraging highly contributing peers tooffer more of their resources.</p> / QC 20100506