Time series forecasting using neural networks vs. Box-Jenkins methodology

Article Abstract:

We discuss the results of a comparative study of the performance of neural networks and conventional methods in forecasting time series. Our work was initially inspired by previously published works that yielded inconsistent results about comparative performance. We have experimented with three time series of different complexity using different feed forward, backpropagation neural network models and the standard Box-Jenkins model. Our experiments demonstrate that for time series with long memory, both methods produced comparable results. However, for series with short memory, neural networks outperformed the Box-Jenkins model. We note that some of the comparable results arise since the neural network and time series model appear to be functionally similar models. We have found that for time series of different complexities there are optimal neural network topologies and parameters that enable them to learn more efficiently. Our initial conclusions are that neural networks are robust and provide good long-term forecasting. They are also parsimonious in their data requirements. Neural networks represent a promising alternative for forecasting, but there are problems determining the optimal topology and parameters for efficient learning. (Reprinted by permission of the publisher.)

Neural networks, Product introduction, Forecasting, Technical, Comparison, Scientific Research, New Technique, Computer Learning, Time Series, Models of Computation, Neural Network

---

Qualitative methodology in simulation model engineering

Article Abstract:

Qualitative methodology in simulation requires high level modeling constructs such as causal graphs, bond graphs, or natural language. Even though there exists a virtual continuum between qualitative and quantitative simulation methods, it is important to discuss qualitative methods since they relate to notions of process abstraction and discovery during model evolution. We will refer to the evolutionary process as simulation model engineering. We define the basic notions of qualitative simulation, give specific examples for a pendulum system, and finally discuss various issues and directions for future work. (Reprinted by permission of the publisher.)

Technology

---

A study of terminology and issues in qualitative simulation

Article Abstract:

Much of the recent AI research related conceptually to simulation may be found under the heading 'qualitative simulation and reasoning'. This paper will present terminology used within AI and critically analyze the terminology with respect to current simulation literature. By discussing terminology, we will help to bridge the gaps between work in qualitative simulation and traditional simulation and provide the foundation for future discussions on the relationships between these two areas. (Reprinted by permission of the publisher.)

Artificial intelligence, Terminology

Similar abstracts:

• Abstracts: Impeller increases mass transfer rates. Vibratory equipment now has smart technology
• Abstracts: Digital Image Generation for a Tank Gunnery Simulator. 'Supercomputer U's'. An Artificially Intelligent Locomotive Mechanic
• Abstracts: Better cars for less. How and when to make tradeoffs
• Abstracts: Power consumers in the dark. The measurement explosion. A powerful time
• Abstracts: Government oversight: fighting waste and fraud. Iraq goes wireless. Radio astronomy: new windows on the universe

This website is not affiliated with document authors or copyright owners. This page is provided for informational purposes only. Unintentional errors are possible.