| There is tremendous research interest in time series-based data mining. Most of these efforts focus upon data representation, pattern identification, rule discovery, and prediction. Discovering accurate and reliable time series algorithmic models is an important challenge to the data mining community. Financial data mining is an obvious domain, which offers both an intellectual and financial appeal to academia and the corporate community. Consequently, the financial domain contains many well-defined and well-understood algorithms, which are typically referred to as technical indicators. Ontological application of these algorithms in another domain is an interesting topic.; Traditionally, data mining focuses on the mining of data, whereas this research adopts an ontological approach by focusing on the mining of algorithms. Specifically, it explores the reusability of well-defined algorithms from a relatively stable domain into a completely different domain. Intuitively, it seems plausible to apply algorithm across different domains.; Two major experiments are conducted which map financial algorithms into the energy domain. Results range from good to outstanding in predicting energy usage based upon temperature. |
