Performance Analysis and Development Potential

Oct 13, 2025 by sorusgentrification.com 85 views

Introduction to Artificial Neural Networks (ANN)

Artificial Neural Networks (ANN) is a technique in artificial intelligence that mimics the workings of nerves in the human brain, processing information obtained. At present, the JST has been used in various real-world problems, including in the fields of health, economy, and agriculture. The JST development process creates various methods, both supervised or unsupervised. This article will discuss the two methods used in the study, namely Fuzzy Logic Sugeno and Backpropagation Levenberg-Marquardt (LM), also compared it with the Bayesian Regularization (BR) backpropagation (BR) method.

Background of the Study

In this study, the data used came from the sale of used Toyota Avanza cars in Medan and surrounding areas. The fuzzy method was chosen because of its ability to handle problems that are uncertain, very suitable for the condition of used cars whose each unit has different characteristics. Fuzzy Logic Sugeno defines data that will be used as input into the backpropagation levenberg-marquardt backpropagation network.

Levenberg-Marquardt Learning Algorithm

Levenberg-Marquardt is a learning algorithm that is known to be efficient in calculations, especially in predicting output in a faster time. In this study, several training and testing with various network architecture were conducted to assess the combination performance of Fuzzy Logic and the Backpropagation method.

Experimental Results

The experimental results showed that the Bayesian Regularization (BR) backpropagation method produced a better Mean Squared Error (MSE) than Levenberg-Marquardt. The MSE produced by BR is 1,164 in the training phase and 1,036 in the testing phase, while Levenberg-Marquardt only produces 10 iterations for its EPOCH and BR produces 79 iterations. This indicates, although LM is faster, the accuracy obtained by the BR method is higher.

Performance Analysis and Development Potential

The performance of the algorithm tested shows that although training with Levenberg-Marquardt is faster in the time needed, the decision to use more complex methods such as Bayesian Regularization still gives more accurate results. This is very important for users in the real-world context, where prediction accuracy has great implications, especially in the automotive industry related to the sale and purchase of used cars in physical and price conditions.

Advantages of Fuzzy Logic Method

The advantage of the Fuzzy Logic method in identifying the status and categorization box of the condition of the car is also an added value in the proposed prediction system. By implementing a fuzzy-based approach, we can confirm that the resulting model can adapt and accommodate uncertain information, which is a common characteristic in the used car market.

Conclusion and Future Research Directions

Thus, this analysis shows the potential for the use of a combination of the JST method that can be more visible in terms of performance. Suggestions for further research are to explore further merging between this method in a broader problematic context and testing in a larger dataset. This will help in strengthening the reliability of predictive models and providing deeper insights about the practical applications of this technology in everyday life.

Limitations and Future Work

That way, this research not only provides a picture of different algorithm performance, but also opens the way for further exploration in this increasingly developed field. The limitations of this study are the use of a limited dataset and the need for further testing in a larger and more diverse dataset. Future work should focus on exploring the application of this method in other fields and developing more complex models that can handle larger and more complex datasets.

Implications of the Study

The implications of this study are significant, as it highlights the potential of using a combination of Fuzzy Logic and Backpropagation methods in predicting the sale price of used cars. This can have significant implications for the automotive industry, as it can help in making more accurate predictions and improving the decision-making process.

References

[1] Levenberg, K. (1944). A method for the solution of certain non-linear problems in least squares. Quarterly of Applied Mathematics, 2(2), 164-168.
[2] Marquardt, D. W. (1963). An algorithm for least-squares estimation of nonlinear parameters. Journal of the Society for Industrial and Applied Mathematics, 11(2), 431-441.
[3] Sugeno, M. (1985). Industrial applications of fuzzy control. Elsevier Science Publishers.
[4] Bayesian Regularization (BR) backpropagation method. (n.d.). Retrieved from https://www.mathworks.com/help/nnet/ref/bayesian-regularization.html

Note: The references provided are a selection of the relevant literature in the field of Artificial Neural Networks and Fuzzy Logic. They are not an exhaustive list and are intended to provide a starting point for further research.

Frequently Asked Questions

Q: What is the main objective of this study?

A: The main objective of this study is to compare the performance of the Backpropagation Levenberg-Marquardt (LM) and Fuzzy Logic Sugeno methods in predicting the sale price of used cars.

Q: What is the significance of using Fuzzy Logic in this study?

A: Fuzzy Logic is used in this study because it can handle problems that are uncertain, which is very suitable for the condition of used cars whose each unit has different characteristics.

Q: What is the Levenberg-Marquardt learning algorithm?

A: Levenberg-Marquardt is a learning algorithm that is known to be efficient in calculations, especially in predicting output in a faster time.

Q: What are the experimental results of this study?

A: The experimental results showed that the Bayesian Regularization (BR) backpropagation method produced a better Mean Squared Error (MSE) than Levenberg-Marquardt.

Q: What are the implications of this study?

A: The implications of this study are significant, as it highlights the potential of using a combination of Fuzzy Logic and Backpropagation methods in predicting the sale price of used cars.

Q: What are the limitations of this study?

A: The limitations of this study are the use of a limited dataset and the need for further testing in a larger and more diverse dataset.

Q: What are the future research directions?

A: Suggestions for further research are to explore further merging between this method in a broader problematic context and testing in a larger dataset.

Q: What are the practical applications of this technology in everyday life?

A: This technology can be used in various fields, including the automotive industry, to make more accurate predictions and improve the decision-making process.

Q: What are the benefits of using a combination of Fuzzy Logic and Backpropagation methods?

A: The benefits of using a combination of Fuzzy Logic and Backpropagation methods include improved accuracy, faster processing time, and the ability to handle uncertain data.

Q: What are the challenges of implementing this technology in real-world applications?

A: The challenges of implementing this technology in real-world applications include the need for a large and diverse dataset, the complexity of the models, and the need for further testing and validation.

Q: What are the future prospects of this technology?

A: The future prospects of this technology are significant, as it has the potential to be used in various fields and to make a significant impact on decision-making processes.

Conclusion

This Q&A article provides a summary of the main points of the study, including the objectives, methods, results, and implications. It also addresses some of the frequently asked questions about the study and provides additional information on the benefits and challenges of using a combination of Fuzzy Logic and Backpropagation methods.

References

[1] Levenberg, K. (1944). A method for the solution of certain non-linear problems in least squares. Quarterly of Applied Mathematics, 2(2), 164-168.
[2] Marquardt, D. W. (1963). An algorithm for least-squares estimation of nonlinear parameters. Journal of the Society for Industrial and Applied Mathematics, 11(2), 431-441.
[3] Sugeno, M. (1985). Industrial applications of fuzzy control. Elsevier Science Publishers.
[4] Bayesian Regularization (BR) backpropagation method. (n.d.). Retrieved from https://www.mathworks.com/help/nnet/ref/bayesian-regularization.html