A comparative evaluation of nature-inspired algorithms for feature selection problems.

Feature selection is a critical component of machine learning and data mining which addresses challenges like irrelevance, noise, redundancy in large-scale data etc., which often result in the curse of dimensionality. This study employs a K-nearest neighbour wrapper to implement feature selection using six nature-inspired algorithms, derived from human behaviour and mammal-inspired techniques. Evaluated on six real-world datasets, the study aims to compare the performance of these algorithms in terms of accuracy, feature count, fitness, convergence and computational cost. The findings underscore the efficacy of the Human Learning Optimization, Poor and Rich Optimization and Grey Wolf Optimizer algorithms across multiple performance metrics. For instance, for mean fitness, Human Learning Optimization outperforms the others, followed by Poor and Rich Optimization and Harmony Search. The study suggests the potential of human-inspired algorithms, particularly Poor and Rich Optimization, in robust feature selection without compromising classification accuracy.

A decision analysis model for material selection using simple ranking process.

A large number of materials and various criteria fashion material selection problems as complex multi-criteria decision-making (MCDM) problems. This paper proposes a new decision-making method called the simple ranking process (SRP) to solve complex material selection problems. The accuracy of the criteria weights has a direct impact on the outcomes of the new method. In contrast to current MCDM methods, the normalization step has been eliminated from the SRP method as a potential source of producing incorrect results. The application of the method is appropriate for situations with high levels of complexity in material selection because it only considers the ranks of alternatives in each criterion. The first scenario of vital-immaterial mediocre method (VIMM) is used as a tool to derive criteria weights based on expert assessment. The result of SRP is compared with a number of MCDM methods. In order to evaluate the findings of analytical comparison, a novel statistical measure known as compromise decision index (CDI) is proposed in this paper. CDI revealed that the MCDM methods' outputs for solving the material selection could not be theoretically proven and requires to be evaluated through practice. As a result, the dependency analysis-an additional innovative statistical measure is introduced to demonstrate the reliability of MCDM methods by assessing its dependency on criteria weights. The findings demonstrated that SRP is extremely reliant on criteria weights and its reliability rises with the number of criteria, making it a perfect tool for solving challenging MCDM problems.

A Simple Drain Current Model for MOS Transistors with the Lorentz Force Effect.

A novel concept of drain current modelling in rectangular normal MOS transistors with the Lorentz force has been proposed for the first time. The single-drain MOS transistor is qualified as a magnetic sensor. To create the Lorentz force, a DC loop current is applied through an on-chip metal loop around the device, and the relation between the applied loop current and the created magnetic field is assumed to be linear in nature. The drain current of the MOS transistor is reduced with the applied Lorentz force from both directions. This change in the drain current is ascribed to a change in mobility in the strong inversion region, and a change in mobility of around 4.45% is observed. To model this change, a set of novel drain current equations, under the Lorentz force, for the strong inversion region has been proposed. A satisfactory agreement of an average error of less than 2% between the measured and the calculated drain currents under the magnetic field created by an on-chip metal loop is achieved.

Drain Current Modulation of a Single Drain MOSFET by Lorentz Force for Magnetic Sensing Application.

This paper reports a detailed analysis of the drain current modulation of a single-drain normal-gate n channel metal-oxide semiconductor field effect transistor (n-MOSFET) under an on-chip magnetic field. A single-drain n-MOSFET has been fabricated and placed in the center of a square-shaped metal loop which generates the on-chip magnetic field. The proposed device designed is much smaller in size with respect to the metal loop, which ensures that the generated magnetic field is approximately uniform. The change of drain current and change of bulk current per micron device width has been measured. The result shows that the difference drain current is about 145 µA for the maximum applied magnetic field. Such changes occur from the applied Lorentz force to push out the carriers from the channel. Based on the drain current difference, the change in effective mobility has been detected up to 4.227%. Furthermore, a detailed investigation reveals that the device behavior is quite different in subthreshold and saturation region. A change of 50.24 µA bulk current has also been measured. Finally, the device has been verified for use as a magnetic sensor with sensitivity 4.084% (29.6 T(-1)), which is very effective as compared to other previously reported works for a single device.