abstract: Learning programming involves a variety of complex cognitive activities, from abstract knowledge construction to structural operations, which include program design,modifying, debugging, and documenting tasks. In this work, the objective was to explore and investigate the barriers and obstacles that programming novice learners encountered and how the learners overcome them. Several lab and classroom studies were designed and conducted, the results showed that novice students had different behavior patterns compared to experienced learners, which indicates obstacles encountered. The studies also proved that proper assistance could help novices find helpful materials to read. However, novices still suffered from the lack of background knowledge and the limited cognitive load while learning, which resulted in challenges in understanding programming related materials, especially code examples. Therefore, I further proposed to use the natural language generator (NLG) to generate code explanations for educational purposes. The natural language generator is designed based on Long Short Term Memory (LSTM), a deep-learning translation model. To establish the model, a data set was collected from Amazon Mechanical Turks (AMT) recording explanations from human experts for programming code lines. To evaluate the model, a pilot study was conducted and proved that the readability of the machine generated (MG) explanation was compatible with human explanations, while its accuracy is still not ideal, especially for complicated code lines. Furthermore, a code-example based learning platform was developed to utilize the explanation generating model in programming teaching. To examine the effect of code example explanations on different learners, two lab-class experiments were conducted separately ii in a programming novices’ class and an advanced students’ class. The experiment result indicated that when learning programming concepts, the MG code explanations significantly improved the learning Predictability for novices compared to control group, and the explanations also extended the novices’ learning time by generating more material to read, which potentially lead to a better learning gain. Besides, a completed correlation model was constructed according to the experiment result to illustrate the connections between different factors and the learning effect. / Dissertation/Thesis / Doctoral Dissertation Engineering 2020

Submitted by Viviane Lima da Cunha (viviane@biblioteca.ufpb.br) on 2015-10-19T10:33:41Z No. of bitstreams: 1 arquivototal.pdf: 1382719 bytes, checksum: 6fb250dd1981beb3c5fd696d07996f1c (MD5) / Rejected by Viviane Lima da Cunha (viviane@biblioteca.ufpb.br), reason: arquivo bloqueado, alterar! on 2015-10-19T10:40:32Z (GMT) / Submitted by Viviane Lima da Cunha (viviane@biblioteca.ufpb.br) on 2015-10-19T10:41:42Z No. of bitstreams: 1 arquivototal.pdf: 1382725 bytes, checksum: 6e60ed8a7122bd2fcb8cd389f8d97261 (MD5) / Approved for entry into archive by Maria Suzana Diniz (msuzanad@hotmail.com) on 2015-10-19T11:07:55Z (GMT) No. of bitstreams: 1 arquivototal.pdf: 1382725 bytes, checksum: 6e60ed8a7122bd2fcb8cd389f8d97261 (MD5) / Approved for entry into archive by Maria Suzana Diniz (msuzanad@hotmail.com) on 2015-10-19T11:09:54Z (GMT) No. of bitstreams: 1 arquivototal.pdf: 1382725 bytes, checksum: 6e60ed8a7122bd2fcb8cd389f8d97261 (MD5) / Made available in DSpace on 2015-10-19T11:10:32Z (GMT). No. of bitstreams: 1 arquivototal.pdf: 1382725 bytes, checksum: 6e60ed8a7122bd2fcb8cd389f8d97261 (MD5) Previous issue date: 2014-12-12 / This study aimed to analyze the influence of the investor sentiment in explaining the returns of anomalies, in Brazilian stock market. Additionally, we analyze whether the price differences caused by investor optimism bias are different from those caused by pessimistic investors. The sample included all companies listed on the BM & FBOVESPA. The data were collected in Economatica®. The calculation of returns were by monthly closing prices were used from June 2000 to June 2014 and from 1999 to 2013 annual financial data. To measure the aggregate investor sentiment index, was considered all issues of shares during the period January 1999 to June 2014 and the volume of trading and securities available in this period. The estimation of investor sentiment, was made use of multivariate technique of Principal Component Analysis to capture the common component in four different proxies for investor behavior. To check how investor sentiment relates to the deficiencies have been empirically tested with the number of returns of the portfolios of Long position, Short and Long-Short 12 anomalies-based strategies; and the sentiment index series built, and its variation from one month to the next. It was found that the measure sentiment index increased explanatory power for much of the anomalies only when included in the CAPM, but by controlling the three-factor model and four factors, the coefficient lost its statistical significance. When using the index change as an explanatory variable, there was a relationship with future returns, robust to all risk factors. Thus, it is possible to relate the investor sentiment index with returns of portfolios formed based on value anomalies. Analyzing the mean returns after periods of optimism and pessimism, there was no statistically significant values sufficient to infer a possible existence of restrictions on sales short, although much of the anomalies present the spread between the average returns after periods optimistic and pessimistic with the expected sign. / Este trabalho teve como objetivo verificar se existe relação entre o sentimento do investidor e as anomalias de mercado no Brasil. Adicionalmente, também foi analisado se os desvios de preços provocados por investidores com viés otimista são diferentes daqueles provocados pelos investidores pessimistas. A população envolveu todas as empresas listadas na BM&FBOVESPA. Os dados utilizados foram coletados no Economatica®. Para cálculo dos retornos, foram utilizados preços de fechamento mensais no período de junho de 2000 a junho de 2014, bem como dados contábeis anuais de 1999 a 2013. Para mensuração do índice de sentimento agregado para o mercado, foram consideradas todas as emissões de ações ocorridas no período de janeiro de 1999 a junho de 2014, bem como o volume de negociações e de títulos disponíveis nesse período. Na estimação do sentimento do investidor, fez-se uso da técnica multivariada de Análise de Componentes Principais, para captar o componente em comum de quatro diferentes proxies para o comportamento do mercado. Para verificar a forma como sentimento do investidor se relaciona com as anomalias, foram testadas empiricamente com a série dos retornos das carteiras de posição Long, Short e Long-Short de 12 estratégias baseadas em anomalias; e com a série do índice de sentimento construído e sua variação de um mês para o outro. Constatou-se que a medida do índice de sentimento apresentou poder explicativo para boa parte das anomalias apenas quando incluída no CAPM, porém ao controlar pelo modelo de três fatores e de quatro fatores, o coeficiente perdeu sua significância estatística. Já na utilização da variação do índice como variável explicativa, observou-se uma relação com os retornos futuros, robustos a todos os fatores de risco. Dessa forma, é possível relacionar o índice de sentimento do investidor com os retornos de carteiras formadas com base em anomalias de valor. Na análise dos retornos médios após os períodos de otimismo e pessimismo, não houve valores estatisticamente significantes suficientes para inferir sobre uma possível existência de restrições às operações de venda a descoberto, apesar de boa parte das anomalias apresentarem o sinal esperado no spread entre a média dos retornos após períodos otimistas e pessimistas.

The thesis two-folded problem aim was to identify and evaluate candidate Machine Learning (ML) methods and performance methods, for predicting the weekly number of covid-19 infections. The two-folded problem aim was created from studying public health studies where several challenges were identified. One challenge identified was the lack of using sophisticated and hybrid ML methods in the public health research area. In this thesis a comparison of ML methods for predicting the number of covid-19 weekly infections has been performed. A dataset taken from the Public Health Agency in Sweden consisting of 101weeks divided into a 60 % training set and a 40% testing set was used in the evaluation. Five candidate ML methods have been investigated in this thesis called Support Vector Regressor (SVR), Long Short Term Memory (LSTM), Gated Recurrent Network (GRU), Bidirectional-LSTM (BI-LSTM) and LSTM-Convolutional Neural Network (LSTM-CNN). These methods have been evaluated based on three performance measurements called Root Mean Squared Error (RMSE), Mean Absolute Error (MAE) and R2. The evaluation of these candidate ML resulted in the LSTM-CNN model performing the best on RMSE, MAE and R2.

Die fondsgebundene Umsetzung einer Long/Short-Strategie stößt schnell an ihre Grenzen, wenn die Regulierungserfordernisse der UCITS IV-Richtlinie 2009/65/EG als rechtlicher Rahmen für den Investmentfonds maßgeblich sind. Die betreffenden Regelungen verlangen einerseits eine diversifizierte Ausrichtung des Portfolios und beschränken das Universum an investierbaren Vermögenswerten auf finanzielle und liquide Produkte. Andererseits führen sie zu einer wesentlichen Begrenzung der zulässigen Anlagetechniken. Die Möglichkeiten zur Hebelinvestition sind streng limitiert und das Durchführen von Leerverkäufen wird vollständig ausgeschlossen. Der Anknüpfungspunkt, die Performance einer Short-Position dennoch in den Fonds zu integrieren, ist die Abkehr von der direkten und physischen Umsetzung hin zu einer indirekten und synthetischen Einbindung, wie sie durch den Einsatz von Derivaten möglich ist. Um die Auswirkungen der Derivate auf das Risiko- und Renditeprofil der Investmentfonds überschaubar und kontrollierbar zu halten, wird die Intensität des Derivatehandels durch das Festsetzen von Risikolimits auf ein vertretbares Maß beschränkt. Die Wahl eines konkreten Derivats beeinflusst die technische Umsetzung der synthetischen Positionsbildung und bestimmt deren assoziierte Vorgaben im Kontext des Risikomanagements. Insofern Derivate bei der Strategieausrichtung des UCITS-Fonds ausgeschlossen werden, lassen sich Short-Positionen lediglich gegenüber aggregierten Exposures in Form einer Dachfondskonstruktion berücksichtigen. Das Ausarbeiten kapitalrechtlicher Vorgaben und das darauf basierende Ableiten von praxisrelevanten Investitionsansätzen, zur Abbildung der Short-Positionen innerhalb einer fondsgebundenen Long/Short-Strategie, stehen im Fokus.

Made available in DSpace on 2016-12-23T14:00:36Z (GMT). No. of bitstreams: 1 Danilo Soares Monte-Mor.pdf: 964390 bytes, checksum: 749870f88ee1c9c692cf782e397379ec (MD5) Previous issue date: 2010-12-17 / Cada vez mais tem surgido no mercado de investimento fundos de retorno absoluto (Hedge Funds) que têm como objetivo principal melhorar seus desempenhos através de estratégias de arbitragem, como é o caso das estratégias long-short. É o comportamento desproporcional e até mesmo antagônico dos preços dos ativos que permite aos players estruturar estratégias para gerar retornos adicionais, superiores aos custos de oportunidade e independentes ao movimento do mercado. Neste trabalho foi utilizada a Teoria de Valores Extremos (TVE), um importante ramo da probabilidade, para que fossem modeladas as séries da relação direta entre preços de dois pares de ativos. Os quantis obtidos a partir de tal modelagem, juntamente com os quantis fornecidos pela normal, foram superpostos aos dados para períodos subsequentes ao período analisado. A partir da comparação desses dados foi criada uma nova estratégia quantitativa long-short de arbitragem, a qual denominamos GEV Long-Short Strategy / Increasingly has appeared on the market of investment Absolute Return Funds (Hedge Funds), which have the main objective to improve their performance through arbitrage strategies, as long-short strategies. It is the disproportionate evolution and even antagonistic of active prices that allows the players to structure strategies to generate additional returns, higher than the opportunity costs and independent of the movement of the market. In this work we used Extreme Value Theory (EVT), an important segment of probability, to model the series of direct relationship between prices of two pairs of assets. The quantiles obtained from such modeling and the quantile provided by normal were superimposed on data for periods subsequent to the period analyzed. From the comparison of such data we created a new quantitative long-short arbitrage strategy, called GEV Long-Short Strategy

Made available in DSpace on 2016-12-23T14:00:36Z (GMT). No. of bitstreams: 1 Danilo Soares Monte-Mor.pdf: 964390 bytes, checksum: 749870f88ee1c9c692cf782e397379ec (MD5) Previous issue date: 2010-12-17 / Increasingly has appeared on the market of investment Absolute Return Funds (Hedge Funds), which have the main objective to improve their performance through arbitrage strategies, as long-short strategies. It is the disproportionate evolution and even antagonistic of active prices that allows the players to structure strategies to generate additional returns, higher than the opportunity costs and independent of the movement of the market. In this work we used Extreme Value Theory (EVT), an important segment of probability, to model the series of direct relationship between prices of two pairs of assets. The quantiles obtained from such modeling and the quantile provided by normal were superimposed on data for periods subsequent to the period analyzed. From the comparison of such data we created a new quantitative long-short arbitrage strategy, called GEV Long-Short Strategy / Cada vez mais tem surgido no mercado de investimento fundos de retorno absoluto (Hedge Funds) que têm como objetivo principal melhorar seus desempenhos através de estratégias de arbitragem, como é o caso das estratégias long-short. É o comportamento desproporcional e até mesmo antagônico dos preços dos ativos que permite aos players estruturar estratégias para gerar retornos adicionais, superiores aos custos de oportunidade e independentes ao movimento do mercado. Neste trabalho foi utilizada a Teoria de Valores Extremos (TVE), um importante ramo da probabilidade, para que fossem modeladas as séries da relação direta entre preços de dois pares de ativos. Os quantis obtidos a partir de tal modelagem, juntamente com os quantis fornecidos pela normal, foram superpostos aos dados para períodos subsequentes ao período analisado. A partir da comparação desses dados foi criada uma nova estratégia quantitativa long-short de arbitragem, a qual denominamos GEV Long-Short Strategy

Företag och organisationer som tillhandahåller kundsupport via e-post kommer över tid att samla på sig stora mängder textuella data. Tack vare kontinuerliga framsteg inom Machine Learning ökar ständigt möjligheterna att dra nytta av tidigare insamlat data för att effektivisera organisationens framtida supporthantering. Syftet med denna studie är att analysera och utvärdera hur Deep Learning kan användas för att automatisera processen att klassificera supportärenden. Studien baseras på ett svenskt företags domän där klassificeringarna sker inom företagets fördefinierade kategorier. För att bygga upp ett dataset extraherades supportärenden inkomna via e-post (par av rubrik och meddelande) från företagets supportdatabas, där samtliga ärenden tillhörde en av nio distinkta kategorier. Utvärderingen gjordes genom att analysera skillnaderna i systemets uppmätta precision då olika metoder för datastädning användes, samt då de neurala nätverken byggdes upp med olika arkitekturer. En avgränsning gjordes att endast undersöka olika typer av Convolutional Neural Networks (CNN) samt Recurrent Neural Networks (RNN) i form av både enkel- och dubbelriktade Long Short Time Memory (LSTM) celler. Resultaten från denna studie visar ingen ökning i precision för någon av de undersökta datastädningsmetoderna. Dock visar resultaten att en begränsning av den använda ordlistan heller inte genererar någon negativ effekt. En begränsning av ordlistan kan fortfarande vara användbar för att minimera andra effekter så som exempelvis träningstiden, och eventuellt även minska risken för överanpassning. Av de undersökta nätverksarkitekturerna presterade CNN bättre än RNN på det använda datasetet. Den mest gynnsamma nätverksarkitekturen var ett nätverk med en konvolution per pipeline som för två olika test-set genererade precisioner på 79,3 respektive 75,4 procent. Resultaten visar också att några kategorier är svårare för nätverket att klassificera än andra, eftersom dessa inte är tillräckligt distinkta från resterande kategorier i datasetet. / Companies and organizations providing customer support via email will over time grow a big corpus of text documents. With advances made in Machine Learning the possibilities to use this data to improve the customer support efficiency is steadily increasing. The aim of this study is to analyze and evaluate the use of Deep Learning methods for automizing the process of classifying support errands. This study is based on a Swedish company’s domain where the classification was made within the company’s predefined categories. A dataset was built by obtaining email support errands (subject and body pairs) from the company’s support database. The dataset consisted of data belonging to one of nine separate categories. The evaluation was done by analyzing the alteration in classification accuracy when using different methods for data cleaning and by using different network architectures. A delimitation was set to only examine the effects by using different combinations of Convolutional Neural Networks (CNN) and Recurrent Neural Networks (RNN) in the shape of both unidirectional and bidirectional Long Short Time Memory (LSTM) cells. The results of this study show no increase in classification accuracy by any of the examined data cleaning methods. However, a feature reduction of the used vocabulary is proven to neither have any negative impact on the accuracy. A feature reduction might still be beneficial to minimize other side effects such as the time required to train a network, and possibly to help prevent overfitting. Among the examined network architectures CNN were proven to outperform RNN on the used dataset. The most accurate network architecture was a single convolutional network which on two different test sets reached classification rates of 79,3 and 75,4 percent respectively. The results also show some categories to be harder to classify than others, due to them not being distinct enough towards the rest of the categories in the dataset.

Die fondsgebundene Umsetzung einer Long/Short-Strategie stößt schnell an ihre Grenzen, wenn die Regulierungserfordernisse der UCITS IV-Richtlinie 2009/65/EG als rechtlicher Rahmen für den Investmentfonds maßgeblich sind. Die betreffenden Regelungen verlangen einerseits eine diversifizierte Ausrichtung des Portfolios und beschränken das Universum an investierbaren Vermögenswerten auf finanzielle und liquide Produkte. Andererseits führen sie zu einer wesentlichen Begrenzung der zulässigen Anlagetechniken. Die Möglichkeiten zur Hebelinvestition sind streng limitiert und das Durchführen von Leerverkäufen wird vollständig ausgeschlossen. Der Anknüpfungspunkt, die Performance einer Short-Position dennoch in den Fonds zu integrieren, ist die Abkehr von der direkten und physischen Umsetzung hin zu einer indirekten und synthetischen Einbindung, wie sie durch den Einsatz von Derivaten möglich ist. Um die Auswirkungen der Derivate auf das Risiko- und Renditeprofil der Investmentfonds überschaubar und kontrollierbar zu halten, wird die Intensität des Derivatehandels durch das Festsetzen von Risikolimits auf ein vertretbares Maß beschränkt. Die Wahl eines konkreten Derivats beeinflusst die technische Umsetzung der synthetischen Positionsbildung und bestimmt deren assoziierte Vorgaben im Kontext des Risikomanagements. Insofern Derivate bei der Strategieausrichtung des UCITS-Fonds ausgeschlossen werden, lassen sich Short-Positionen lediglich gegenüber aggregierten Exposures in Form einer Dachfondskonstruktion berücksichtigen. Das Ausarbeiten kapitalrechtlicher Vorgaben und das darauf basierende Ableiten von praxisrelevanten Investitionsansätzen, zur Abbildung der Short-Positionen innerhalb einer fondsgebundenen Long/Short-Strategie, stehen im Fokus.

No description available.

LSTM neural networks have been used for speech recognition, image recognition and other artificial intelligence applications for many years. Most applications perform the LSTM algorithm and the required calculations on cloud computers. Off-line solutions include the use of FPGAs and GPUs but the most promising solutions include ASIC accelerators designed for this purpose only. This report presents an ASIC design capable of performing the multiple iterations of the LSTM algorithm on a unidirectional and without peepholes neural network architecture. The proposed design provides arithmetic level parallelism options as blocks are instantiated based on parameters. The internal structure of the design implements pipelined, parallel or serial solutions depending on which is optimal in every case. The implications concerning these decisions are discussed in detail in the report. The design process is described in detail and the evaluation of the design is also presented to measure accuracy and error of the design output.This thesis work resulted in a complete synthesizable ASIC design implementing an LSTM layer, a Fully Connected layer and a Softmax layer which can perform classification of data based on trained weight matrices and bias vectors. The design primarily uses 16-bit fixed point format with 5 integer and 11 fractional bits but increased precision representations are used in some blocks to reduce error output. Additionally, a verification environment has also been designed and is capable of performing simulations, evaluating the design output by comparing it with results produced from performing the same operations with 64-bit floating point precision on a SystemVerilog testbench and measuring the encountered error. The results concerning the accuracy and the design output error margin are presented in this thesis report. The design went through Logic and Physical synthesis and successfully resulted in a functional netlist for every tested configuration. Timing, area and power measurements on the generated netlists of various configurations of the design show consistency and are reported in this report. / LSTM neurala nätverk har använts för taligenkänning, bildigenkänning och andra artificiella intelligensapplikationer i många år. De flesta applikationer utför LSTM-algoritmen och de nödvändiga beräkningarna i digitala moln. Offline lösningar inkluderar användningen av FPGA och GPU men de mest lovande lösningarna inkluderar ASIC-acceleratorer utformade för endast dettaändamål. Denna rapport presenterar en ASIC-design som kan utföra multipla iterationer av LSTM-algoritmen på en enkelriktad neural nätverksarkitetur utan peepholes. Den föreslagna designed ger aritmetrisk nivå-parallellismalternativ som block som är instansierat baserat på parametrar. Designens inre konstruktion implementerar pipelinerade, parallella, eller seriella lösningar beroende på vilket anternativ som är optimalt till alla fall. Konsekvenserna för dessa beslut diskuteras i detalj i rapporten. Designprocessen beskrivs i detalj och utvärderingen av designen presenteras också för att mäta noggrannheten och felmarginal i designutgången. Resultatet av arbetet från denna rapport är en fullständig syntetiserbar ASIC design som har implementerat ett LSTM-lager, ett fullständigt anslutet lager och ett Softmax-lager som kan utföra klassificering av data baserat på tränade viktmatriser och biasvektorer. Designen använder huvudsakligen 16bitars fast flytpunktsformat med 5 heltal och 11 fraktions bitar men ökade precisionsrepresentationer används i vissa block för att minska felmarginal. Till detta har även en verifieringsmiljö utformats som kan utföra simuleringar, utvärdera designresultatet genom att jämföra det med resultatet som produceras från att utföra samma operationer med 64-bitars flytpunktsprecision på en SystemVerilog testbänk och mäta uppstådda felmarginal. Resultaten avseende noggrannheten och designutgångens felmarginal presenteras i denna rapport.Designen gick genom Logisk och Fysisk syntes och framgångsrikt resulterade i en funktionell nätlista för varje testad konfiguration. Timing, area och effektmätningar på den genererade nätlistorna av olika konfigurationer av designen visar konsistens och rapporteras i denna rapport.