We previously introduced the generalized Weighted Relevance Aggregation Operators (WRAO) for hierarchical fuzzy signatures. WRAO enhances the ability of the fuzzy signature model to adapt to different applications and simplifies the learning of fuzzy signature models from data. In this paper we overcome the practical issues which occur when learning WRAO from data. This paper discuss an algorithm for learning WRAO using the LevenbergMarquardt (LM) method, which is one of the most sophisticated and widely used gradient based optimization method. Also, this paper shows the successful results of applying the proposed algorithm to extract WRAO for two real world problems namely High Salary Selection and SARS Patient Classification.
The rapid adoption of the IEEE 802.11g wireless LAN (WLAN) standard has created a demand for low-cost, small-form-factor implementations. An integrated SoC that implements all of the functions of an 802.11g WLAN in a 0.18µm CMOS technology is presented in this paper. As shown in Fig. 5.2.1, the SoC consists of a 2.4GHz RF transceiver, analog baseband filters, data converters, digital physical layer (PHY), and media access controller (MAC). This IC essentially connects the RF antenna to the digital host computer. The one-chip solution reduces overall package cost and form factor. In addition, integration eliminates the area and power associated with driving package pins in a multi-chip implementation [1,2]. Furthermore, the merging of the analog and digital blocks on the same chip enables a wide digital-analog interface that allows for the use of sophisticated digital signal processing techniques to mitigate analog and RF impairments [2,3].Figure 5.2.2 shows the block diagram of the RF transceiver. A sliding-IF dual-conversion architecture is employed for both the receiver and transmitter. The RF local oscillator (LO RF ) frequency at 2/3 f RF (1.6GHz) and IF local oscillator (LO IF ) frequency at 1/3 f RF (0.8GHz) are generated from an integer-N synthesizer operating at 3.2GHz (twice the LO RF frequency). The voltage-controlled oscillator (VCO) is a varactor-tuned, inductively loaded LC tank. The choice of a VCO frequency at 2⋅LO RF reduces the size of the on-chip spiral inductor. The VCO output is divided by two and four to generate LO RF and LO IF . Inductive tuning is not used in the high-speed dividers because of the high f T available in the 0.18µm process. The loop filter is integrated on-chip using poly resistors and NMOS device capacitors. The synthesizer phase noise measured at the transmitter output, shown in Fig. 5.2.3, is -105dBc/Hz at 100kHz offset.The receiver in Fig. 5.2.2 converts the incoming RF signal to quadrature baseband outputs, while providing enough gain and interference rejection. The frequency allocation results in approximately 1.6GHz (2/3 f RF ) frequency separation between the incoming RF signal and the image channel. As a result, the bandpass elements used in the RF stages provide adequate image rejection, thereby eliminating the need for an explicit image rejection filter. The low-noise amplifier (LNA) consists of a cascoded differential pair with inductive degeneration and loading. The LNA has an adjustable RF gain to accommodate large RF input signals. In the low gain setting, 75% of the bias current is diverted from the output load to the supply, which results in a 12dB gain reduction. Figure 5.2.4 shows a circuit diagram of the dual down-conversion mixers. The RF mixer down-converts the 2.4GHz signal to an IF frequency of 800MHz. NMOS diode loads, which improve the IF gain stability over process and temperature, are used instead of bulky inductors. The NMOS loads are placed in deep N-wells to allow a source-to-bulk connection that removes transistor body effects. A common-mode f...
Animal models of tinnitus rely on interpretation of behavioural or reflexive tests to determine the presence of this phantom perception. A commonly used test is the gap prepulse inhibition of acoustic startle (GPIAS), which is often combined with prepulse inhibition (PPI) to ensure that reduced GPIAS suppression is not due to hearing loss caused by the acoustic trauma commonly used to trigger tinnitus development. In our laboratory GPIAS and PPI are routinely used on two colonies of outbred tri-colour guinea pigs. However, our results show that these colonies show divergent results even before any tinnitus-inducing treatment, which impacts their suitability in tinnitus models. Although colony 1 and 2 show similar results in PPI (~95% of animals showing significant suppression), only ~30% of colony 2 also shows significant suppression in GPIAS compared to ~75% of colony 1. Cochlear sensitivity measured using compound action potentials showed no significant differences between colonies. Therefore, peripheral threshold loss was excluded as a possible factor. Our results show that similar strains of laboratory animals can show highly divergent results and GPIAS testing for tinnitus will not work for every animal strain. In addition, our data support the notion that PPI and GPIAS responses may rely on different neural circuitry.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.
hi@scite.ai
10624 S. Eastern Ave., Ste. A-614
Henderson, NV 89052, USA
Copyright © 2024 scite LLC. All rights reserved.
Made with 💙 for researchers
Part of the Research Solutions Family.