University of Twente Student Theses
Predicting race results using artificial neural networks
Stoppels, Eloy (2017) Predicting race results using artificial neural networks.
|
PDF
2MB |
| Abstract: | In this thesis Artificial Neural Networks are used to predict Formula One finish results. The last four races of the season 16/17 are predicted based on the first seventeen races. The first part of this thesis is dedicated to the theory behind Artificial Neural Networks. The aim is to give readers insights in the world and terminology of artificial neural networks. All the important terminology is discussed and explained, using some simple examples. Furthermore some key aspect are discussed in more depth. We look at what the in uences of multiple layers and multiple neurons are. We make clear why a deep neural network is used. The next key as- pect we discuss is; 'training an artificial neural network'. During training the free parameters are optimized. This optimization is done by solving a minimization problem. Therefore, training an artificial neural network can be done by using state-of-the-art optimization methods. Other key aspect are the different activation functions and cost functions which are used in artificial neural networks. Lastly, we explain how possible errors during training can be avoided by using so called regularization methods. The second part is dedicated to the experimental research. First we give and explain public available data which we use as prediction features. Afterwards we show how to initialize an arti- ficial neural network, where we show how multiple layers, activation functions, etc. in uence the predicted results. We finish the experimental research by using the initialized network to predict the outcome of the last four Formula One races of the season 16/17. We compare three different data-sets, the first is the actual data of the season 16/17, the other two are data-sets where we added data in a certain way. Next to this, we compare the predicted results of the artificial neural network with two simple prediction methods and with multiclass logistic regression. The conclu- sion of this comparison is that using artificial neural networks has benefits as it predicts better outcome results. Keywords: Artificial Neural Networks, multi-layer network, activation function, cost function, regularization, minimization, optimization |
| Item Type: | Essay (Master) |
| Faculty: | EEMCS: Electrical Engineering, Mathematics and Computer Science |
| Subject: | 31 mathematics |
| Programme: | Applied Mathematics MSc (60348) |
| Link to this item: | https://purl.utwente.nl/essays/74765 |
| Export this item as: | BibTeX EndNote HTML Citation Reference Manager |
Show download statistics for this publication
Show download statistics for this publicationRepository Staff Only: item control page