Does anyone use Fuzzy to combine deep learning? I would like to ask what is the deep fuzzy model in deep learning for time series problem? Have the some source code (python/Java) to forecasting the time series data set? Thank you.
Yu-Da,
Fuzzy logic was introduced by Lotfi A. Zadeh in 1965. Fuzzy logic is primarily based on the idea of fuzzy sets, which could be utilized to model linguistic terms (i.e. human expressions such as small, large, etc.). “In fuzzy
logic, it is possible to formulate fuzzy rules that incorporate linguistic expressions, and apply the rules to decision-making processes.” Similar to neural networks, fuzzy systems are utilized for a wide range of purposes from control functions to data analysis and decision making problems.
Fuzzy systems can be used for the same purposes as neural networks. The difference between the two is that fuzzy systems are constructed on explicit information, which is ultimately expressed in the form of linguistic rules rather than being built by a learning algorithm. Due to this difference, applying learning algorithms to a fuzzy system presents its difficulties. However, because
both neural networks and fuzzy systems became popular during the late 1980s, the two were combined to form neuro-fuzzy systems.
I would look further into Adaptive Neuro-Fuzzy Inference Systems (ANFIS).
Hope this helps.
Thank you, Niaki.
I know the ANFIS which apply fuzzy theory into NN model. However, the recently CFP on Transactions on Fuzzy Systems : Special Issue on Deep Fuzzy Models. I google the deep fuzzy models, and there are not information about deep fuzzy model. I consider this may be a deep NN based on fuzzy.
- Badges
- Science topic
- Similar topics
- Computer Science and Engineering
- Programming Languages
- Python