Artificial Intelligence in Progress

Home   Artificial Intell.   Bio-exact   Chemistry   Computer Sci. BSc   Earth Sciences   Mathematics   Physics & Astr.   Science Educ.   Submit   Webmaster        
bachelors   masters   all  


Displaying theses 1-10 of 440 total
Previous  1  2  3  4  5  6  7  8  9  10  11  Next  Search:

J.W.H. Verdegaal
Master programme: Artificial Intelligence June 21st, 2018
Institute: ILLC Research group: Language and Computation Graduation thesis Supervisor: R. Fernandez Rovira
External Memory Enhanced Sequence-to-Sequence Dialogue Systems
The sequence-to-sequence model has proven itself in machine translation. Since dialogue is technically quite similar to translation, the same model can be used to generate replies. Although, the similarity only holds for the ability to generate grammatical and coherent sentences or replies. When the context size (number of utterances used as input for the dialogue system) grows, the attention mechanism is hypothesized to be too simple to be able to reason about content in order to generate a suitable reply. External memory can replace attention in the sequence-to-sequence model and the ReasoningNSE (Neural Semantic Encoder) and DNC (Differentiable Neural Computer) are tested and compared over a growing context size.
picture that illustrates the research done
Scientific abstract (pdf 1K)   Full text (pdf 1611K)

M. Winkels
Master programme: Artificial Intelligence April 12th, 2018
Institute: Informatics Institute Research group: Amsterdam Machine Learning Lab Graduation thesis Supervisor: Taco Cohen
Data-efficient learning for pulmonary nodule detection
Convolutional neural networks -- the methodology of choice for automated image analysis -- typically require a large amount of annotated data to learn from, which is difficult to obtain in the medical domain. This work shows that the sample complexity for automated medical image analysis tasks can be significantly improved by using 3D roto-translation group convolutions instead of more conventional translational convolutions. 3D G-CNNs were applied to the false positive reduction step of pulmonary nodule detection, and proved to be substantially more effective in terms of performance, sensitivity to malignant nodules, and speed of convergence compared to a baseline architecture with regular convolutions and a similar number of parameters. For every dataset size, the G-CNNs substantially outperformed the baseline CNN, achieving scores very close to or exceeding those of the CNN trained on ten times more data, with differences in performance being more pronounced in the small-data regime.
picture that illustrates the research done
Scientific abstract (pdf 1K)   Full text (pdf 7942K)

R.R. van der Heijden
Master programme: Artificial Intelligence February 5th, 2018
Institute: VU / Other Research group: Knowledge Representation and Reasoning Graduation thesis Supervisor: Maarten van Someren
Modeling a Diabetes Guideline using the TMR model for interaction detection
Clinical guidelines are used to standardize medical treatments and help doctors with best practices in their field. The TMR model is used to detect conflicts between different guidelines. We have performed a case study of a real world diabetes guideline using the TMR model. We found 51 conflicts between 21 selected recommendations and discuss their implications. We also extend the TMR model to be able to handle preconditions of recommendations. With our extension we are able determine if the preconditions of a supposed conflict prevent a real conflict from occurring. Our results indicate that 3 of the 51 detected conflicts can indeed be ruled out because of this.
picture that illustrates the research done
Scientific abstract (pdf 1K)   Full text (pdf 566K)

M. Wardenaar
Master programme: Artificial Intelligence January 12th, 2018
Institute: Informatics Institute Research group: Information and Language Processing Systems Graduation thesis Supervisor: Ilya Markov
A Non-Sequential Neural Click Model for Web Search
Getting a better understanding of click behavior is important for advancing information retrieval systems. The aim of this work is to create a click model that not only predicts what documents of a Search Engine Result Page (SERP) the user will click on, but also in what order. To do this two Deep Learning models are implemented and compared to the recently presented Neural Click Model. The Neural Click Model (Borisov et al., 2016) was recently introduced and managed to outperform existing click models that rely on the Probabilistic Graphical Model (PGM) framework. One of the models introduced in this research uses an attention mechanism to give more information about the documents at the time of predicting the clicks. The results show that the non sequential neural click models are a promising subject for further research. The results also show that an attention mechanism has a positive effect on the performance of the click model. Another contribution of this research is to introduce a method for evaluating the performance of click models on non sequential click sequences.
picture that illustrates the research done
Scientific abstract (pdf 149K)   Full text (pdf 629K)

K. Bouwens
Master programme: Artificial Intelligence January 5th, 2018
Institute: Informatics Institute Research group: Amsterdam Machine Learning Lab Graduation thesis Supervisor: Peter O'Connor
photo of the author
Latent Customer Representations for Credit Card Fraud Detection
This thesis is about the detection of fraudulent credit card transactions. In order to do this, we have applied a type of probabilistic graphical model called Bayesian Networks. Moreover we aim to create latent representations of different aspects of transactions. This is done by modeling fraudulent transactions, non-fraudulent transactions and customers. We show that our models outperform some state-of-the-art cost-sensitive models. And that we are able to do this without using elaborate feature engineering strategies.
picture that illustrates the research done
Scientific abstract (pdf 2K)   Full text (pdf 389K)

S. Negrijn
Master programme: Artificial Intelligence December 19th, 2017
Institute: Informatics Institute Research group: Computer Vision Graduation thesis Supervisor: Dr T.E.J. Mensink
Exploiting Symmetries to Relocalise in RoboCup Soccer
In this thesis we are interested in exploiting symmetries in order to more accurately determine the location of a robot while playing robotic soccer. For many autonomous robots, localisation is an essential task as without it the next action can not reliably be determined. For example, a robot can not run clear if it has no understanding of its own position in the field, the position of the ball and the position of its teammates and opponents. When the robot no longer knows where it is at it needs to redetermine, or relocalise, its location without being able to use its previously known position. By exploiting the perfect symmetries of a soccer field and extending PoseNet, a relocalisation model, we use a single color image to predict the position of the robot. Instead of generating these images with a real robot, a simulator is created to easily be able to change conditions such as the lighting, field colour and even the surroundings of the field. After showing that using these symmetries in a soccer field increases the accuracy of the location predictions, we also show an increase when compared to the original PoseNet architecture and its Street Scenes data sets.
picture that illustrates the research done
Scientific abstract (pdf 28K)   Full text (pdf 5795K)

C. Zavou
Master programme: Artificial Intelligence December 18th, 2017
Institute: UvA / Other Research group: UvA / Other Graduation thesis Supervisor: dr. Maarten de Rijke
Question Retrieval in Community Question Answering Enhanced by Tags Information in a Deep Neural Network Framework
Community Question Answering (CQA) platforms need to be easy and fast in question or answer exploration. It is common to use tags to categorize items in these platforms, and create taxonomies that assist exploration, indexing and searching. The focus of this thesis lies in recommending similar questions (Question Retrieval) by simultaneously deciding whether the contexts of two questions are similar, and which tags are applicable for each question. Current methods targeted for Question Retrieval in CQA either consider deep learning approaches (Lei et. al. 2016, Bogdanova et al. 2015), or conventional approaches that utilize the available information on questions' tags (Cao et. al. 2010, Zhou et al. 2013). The former framework is proved to be more powerful –especially in the case with loads of available data–, while the later is faster and successful in cases with few data. In this thesis, a deep learning approach for both question retrieval and tags recommendation is proposed, and their joint learning is found successful for transferring knowledge in the Question Retrieval task, after applying it on the AskUbuntu forum data. Additionally the neural network based Tag Recommendation performs better than the existing conventional methods.
picture that illustrates the research done
Scientific abstract (pdf 2K)   Full text (pdf 1281K)

M.J. Scheepers
Master programme: Artificial Intelligence December 14th, 2017
Institute: Informatics Institute Research group: Information and Language Processing Systems Graduation thesis Supervisor: Evangelos Kanoulas
Improving the Compositionality of Word Embeddings
Word embedddings are mathematical representations of word meaning. Combining these representations allows us to create meaning for statements, sentences or even entire paragraphs. In this thesis we attempt to find better word embeddings which are easier to combine, i.e. compose. Our results indicate that we can improve the compositionality for four popular and widely used word embeddings.
picture that illustrates the research done
Scientific abstract (pdf 79K)   Full text (pdf 2303K)

K.T. Keune
Master programme: Artificial Intelligence December 13th, 2017
Institute: UvA / Other Research group: UvA / Other Graduation thesis Supervisor: Maarten van Someren
Robust Detection of Anomaly Types in Energy Consumption Data
Improving the energy consumption in buildings is an ongoing problem. One of the ways to improve the energy consumption in buildings is by detecting anomalies. This thesis makes a clear distinction between anomaly types, while other approaches do not. This paper introduces different anomaly types based on the anomalies that occur in energy consumption data and presents a robust method to detect them. The anomaly types occur in the context of the outside temperature and the time in hours. They are detected with a robust method that consists of finding a regression model of the data with the context and an anomaly detection rule based on the robust residuals of the regression model. The results show that the general anomaly detection approach could be used for the detection of anomaly types in energy consumption data.
picture that illustrates the research done
Scientific abstract (pdf 2K)   Full text (pdf 336K)

L. Mooiman
Master programme: Artificial Intelligence November 28th, 2017
Institute: UvA / Other Research group: UvA / Other Graduation thesis Supervisor: Maarten de Rijke
Category suggestion for e-commerce queries
This thesis proposes a category suggestion model for the e-commerce queries from the Dutch company Bol.com . The model provides insight into the intent of the visitors and consists of two parts. Firstly, the model clusters the queries by various semantic or lexical similarities. Secondly, a category is assigned to each cluster of queries by looking at the number of clicks on the items after a query is submitted. This results in multiple queries in various clusters with an associated category distribution. An newly submitted query needs to inherit one of these category distributions if you want to provide a category to an unseen query. This is done by using the same lexical and semantic similarities. However, an unseen query is compared to all the query clusters instead of comparing every query. The unseen query inherits the category from the most similar cluster. The final results vary depending on the chosen similarities when clustering the queries and when assigning an unseen query to a cluster. In general, the model can suggest correct categories for a query and provides an insight into the visitors' intent.
picture that illustrates the research done
Scientific abstract (pdf 2K)   For more info or full text, mail to: m.derijke@uva.nl

Previous  1  2  3  4  5  6  7  8  9  10  11  Next  

This page is maintained by thesis@science.uva.nl