Browsing by Author "Leandro, Carlos"
Now showing 1 - 2 of 2
Results Per Page
Sort Options
- Symbolic Knowledge Extraction from Trained Neural Networks Governed by Lukasiewicz LogicsPublication . Leandro, Carlos; Pinheiro Pita, Helder Jorge; Monteiro, LuísThis work describes a methodology to extract symbolic rules from trained neural networks. In our approach, patterns on the network are codified using formulas on a Lukasiewicz logic. For this we take advantage of the fact that every connective in this multi-valued logic can be evaluated by a neuron in an artificial network having, by activation function the identity truncated to zero and one. This fact simplifies symbolic rule extraction and allows the easy injection of formulas into a network architecture. We trained this type of neural network using a back-propagation algorithm based on Levenderg-Marquardt algorithm, where in each learning iteration, we restricted the knowledge dissemination in the network structure. This makes the descriptive power of produced neural networks similar to the descriptive power of Lukasiewicz logic language, minimizing the information loss on the translation between connectionist and symbolic structures. To avoid redundance on the generated network, the method simplifies them in a pruning phase, using the "Optimal Brain Surgeon" algorithm. We tested this method on the task of finding the formula used on the generation of a given truth table. For real data tests, we selected the Mushrooms data set, available on the UCI Machine Learning Repository.
- Wind power forecasting with machine learning: single and combined methodsPublication . Rosa, J.; Pestana, Rui; Leandro, Carlos; Brás-Geraldes, Carlos; Esteves, João; Carvalho, D.In Portugal, wind power represents one of the largest renewable sources of energy in the national energy mix. The investment in wind power started several decades ago and is still on the roadmap of political and industrial players. One example is that by 2030 it is estimated that wind power is going to represent up to 35% of renewable energy production in Portugal. With the growth of the installed wind capacity, the development of methods to forecast the amount of energy generated becomes increasingly necessary. Historically, Numerical Weather Prediction (NWP) models were used. However, forecasting accuracy depends on many variables such as on-site conditions, surrounding terrain relief, local meteorology, etc. Thus, it becomes a challenge to obtain improved results using such methods. This article aims to report the development of a machine learning pipeline with the objective of improving the forecasting capability of the NWP’s to obtain an error lower than 10%.