Saudi Cultural Missions Theses & Dissertations
Permanent URI for this communityhttps://drepo.sdl.edu.sa/handle/20.500.14154/10
Browse
2 results
Search Results
Item Restricted Improving Model-fitting for Computational Models of Neural Function and Behaviour(King's College London, 2024) Ramis, Sarah; wise, TobyParameter estimation plays a critical role in computational neuroscience, enabling the design of models that can predict neural mechanisms under various conditions using statistical methods. However, accurate estimation becomes challenging when the likelihood function is intractable due to the model’s complexity and high dimensionality. With the advent of artificial intelligence, various machine learning algorithms have been developed to estimate the posterior distribution of observed data using simulation-based inference approaches. This study aims to elucidate the capabilities of these machine learning models compared to conventional methods to aid in model selection. Rescorla-Wagner simulated data is used to compare the performance of several techniques: Markov Chain Monte Carlo (MCMC) as a conventional method, Automatic Posterior Transformation (APT) using the Masked Auto-regressive Flow model, and APT enhanced with Gated Recurrent Units (APT-GRU) for data embedding, to estimate the parameters alpha and beta. The experiment involved 12 trials with varying parameter pairs and data volumes. The models were evaluated based on computational time, Mean Squared Error (MSE), Mean Absolute Error (MAE), a customized fit metric, Kullback-Leibler (KL) divergence, convergence, and log-likelihood. Results indicated that computational time increased across all models as dataset volume grew, with MCMC requiring significantly more time than the others. APT-GRU achieved lower MSE and MAE scores than the other models, indicating higher accuracy, followed by APT. Both APT-based models demonstrated stable convergence on large datasets, with APT-GRU showing enhanced convergence on smaller datasets. APT models also approached zero in KL divergence scores and positive log-likelihood values, signifying a good model fit. MCMC, while performing better in estimating alpha than beta, showed large negative KL divergence values, suggesting underfitting. Furthermore, MCMC exhibited lower performance stability across trials, with higher estimation variances. In summary, the findings highlight the strengths of APT and APT-GRU in parameter estimation within the Rescorla-Wagner framework. They suggest that the inclusion of GRU for data embedding improves both model robustness and estimation accuracy. These insights pave the way for future research into exploring different machine learning algorithms and model architectures for parameter estimations.12 0Item Restricted A Comparison of Time Series and Deep Learning Methods for Predicting Stock Prices(Saudi Digital Library, 2023-03-23) Alajmi, Shahad; Ji, LanpengStock market is one of the most competitive financial markets where investors need to know the trend of prices in advance. There have been many improvements and advancements in the application of neural networks in the financial industry. In this research, two advanced methods were used to simulate and predict the close stock prices of Saudi Telecom Company (STC). The first method was the autoregressive integrated moving average (ARIMA) and the second method was a by using a class of deep learning neural networks called recurrent neural networks (RNN). ARIMA (p,d,q) was the statistical method selected as a time series model based on the level, trend ,and seasonality of data. Additionally, the order of p and q is based on an autocorrelation function ‘ACF’ ,and a partial autocorrelation function ‘PACF’, the optimal model of this research was ARIMA (1,0,28). Moreover, RNN uses long short-term memory layers (LSTMs), dropout regularisation, activation functions, a loss function, which is the mean square error (MSE), and the Adam optimiser to simulate the predictions. The unique characteristic of LSTMs is that the model is able to store previous data over time and use this data to predict future prices. The structure of the LSTM consists of five layers: one input layer, three hidden layers ,and one output layer. The methods used to measure the performance of predictions of each model are the mean absolute percentage error ‘MAPE’ and the root squared mean error ‘RMSE’. ARIMA (1,0,28) is the model that was found to have a lower error between actual and predicted prices. After analysing each model, ARIMA model’s prediction accuracy was 96.3% and RNN’s accuracy was 93.8%; we concluded that the ARIMA model is better than the RNN model for forecasting the close stock prices of STC. This research include important data which can benefit investors and companies to make economical decisions, such as when to buy or sell shares.26 0