In data mining, objects are often represented by a set of features, where each feature of an object has only one value. However, in reality, some features can take on multiple values, for instance, a person with several job titles, hobbies, and email addresses. These features can be referred to as set-valued features and are often treated with dummy features when using existing data mining algorithms to analyze data with set-valued features. In this paper, we propose an SV- $k$ -modes algorithm that clusters categorical data with set-valued features. In this algorithm, a distance function is defined between two objects with set-valued features, and a set-valued mode representation of cluster centers is proposed. We develop a heuristic method to update cluster centers in the iterative clustering process and an initialization algorithm to select the initial cluster centers. The convergence and complexity of the SV- $k$ -modes algorithm are analyzed. Experiments are conducted on both synthetic data and real data from five different applications. The experimental results have shown that the SV- $k$ -modes algorithm performs better when clustering real data than do three other categorical clustering algorithms and that the algorithm is scalable to large data.

No abstract

No abstract

Batch-type hot rolling planning highly affects electricity costs in a steel plant, but previous research models seldom considered time-of-use (TOU) electricity pricing. Based on an analysis of the hot-rolling process and TOU electricity pricing, a batch-processing plan optimization model for hot rolling was established, using an objective function with the goal of minimizing the total penalty incurred by the differences in width, thickness, and hardness among adjacent slabs, as well as the electricity cost of the rolling process. A method was provided to solve the model through improved genetic algorithm. An analysis of the batch processing of the hot rolling of 240 slabs of different sizes at a steel plant proved the effectiveness of the proposed model. Compared to the man–machine interaction model and the model in which TOU electricity pricing was not considered, the batch-processing model that included TOU electricity pricing produced significantly better results with respect to both product quality and power consumption.

This study presents the room-temperature operation of deep-trench type nitride–oxide metal–insulator–semiconductor three-terminal tunneling devices which were fabricated by a standard metal–oxide–semiconductor process. It is instructive to observe a photoinduced N-type negative differential resistance (NDR) with a high peak-to-valley current ratio for device operated under negative polysilicon node bias under tungsten lamp illumination. An explanation was provided for the NDR phenomenon with proper three-terminal biasing. The sudden current drop under light illumination was caused by the sudden reduction of the two-carrier conduction due to Esaki band-to-band tunneling. The NDR amplitude could be modulated by light intensity. The position of the NDR current peak was tunable at different voltages with different p-well biases. The optoelectronic response of nitride–oxide devices we investigate here may open an application window for the nitride–oxide system in silicon-based optoelectronic integrated circuits, wireless communications, and future quantum devices.