Abstract:
Techniques for generating machine learning architectures are provided. A data set is received for training one or more machine learning (ML) models, where the data set comprises labeled exemplars for a plurality of classes. The data set is partitioned into a training set and a testing set. A first ML model is trained using the training set, and a quality of the first ML model with respect to each class of the plurality of classes is evaluated using the testing set. Upon determining that the quality of the first ML model is below a predefined threshold with respect to a first class and a second class of the plurality of classes, a subset of the training set is identified, where each exemplar in the subset corresponds to either the first class or the second class. A second ML model is trained using the subset of the training set.
Abstract:
A dynamic pulse width modulation (PWM) amplifier comprising an input terminal, a dynamic PWM controller, a power stage, a low pass filter, and an output terminal. The input terminal receives an input signal. The dynamic PWM controller transforms the N-bit input signal sampled by a sampling frequency to a 1-bit PWM signal. The power stage receives and outputs the 1-bit PWM signal. The low pass filter receives and outputs the 1-bit PWM signal. The 1-bit PWM signal is used to drive the power stage and the low pass filter. The output terminal outputs the 1-bit PWM signal. The dynamic PWM amplifier is characterized in the dynamic PWM amplifier uses a register array to store the input signal processed immediately in each frame, regroups the input signal, and outputs the regroup signal, so that the 1-bit PWM signal is changed according to the input signal.
Abstract:
A semiconductor structure is disclosed, including a substrate having therein a first well of a first conductivity type and a second well of a second conductivity type, a first MOS transistor of the first conductivity type and a second MOS transistor of the second conductivity type. The first MOS transistor is disposed on the second well, including a gate structure on the second well and a strained layer of the first conductivity type in an opening in the second well beside the gate structure. The difference between the lattice parameter of a portion of the strained layer near the bottom of the opening and that of the substrate is less than the difference between the lattice parameter of a portion of the strained layer apart from the bottom of the opening and that of the substrate. The second MOS transistor is disposed on the first well.
Abstract:
Highly crimped, fully drawn bicomponent fibers, prepared by melt-spinning, followed by gas-flow quenching, heat treatment and high speed windup, are provided, as are fine-decitex and highly uniform polyester bicomponent fibers.
Abstract:
The present invention relates to a radiation measurement instrument calibration facility with the abilities of lowering scattered radiation and shielding background radiation and it is capable of providing a suitable environment for performing performance test, calibration and experiment upon a radiation measurement instrument. In an embodiment, the calibration facility comprises: a shielding device, a collimator, a multi-source irradiator, a radiation baffle, a carrier, an electric door unit and a control unit. With the design of the calibration facility of the present invention, the interference coming from the background radiation and scattered radiation in the laboratory during the radiation measurement instrument calibration can be effectively reduced to enhance the accuracy of measurement or calibration for the instrument, and also the instrument calibration and testing can be performed in radiation fields of low-, medium- and high-dose rate levels to meet the requirements of ISO 4037-1 (1996) Standard.
Abstract:
A semiconductor structure is disclosed, including a substrate having therein a first well of a first conductivity type and a second well of a second conductivity type, a first MOS transistor of the first conductivity type and a second MOS transistor of the second conductivity type. The first MOS transistor is disposed on the second well, including a gate structure on the second well and a strained layer of the first conductivity type in an opening in the second well beside the gate structure. The difference between the cell parameter of a portion of the strained layer near the bottom of the opening and that of the substrate is less than the difference between the cell parameter of a portion of the strained layer apart from the bottom of the opening and that of the substrate. The second MOS transistor is disposed on the first well.
Abstract:
A semiconductor structure is disclosed, including a substrate having therein a first well of a first conductivity type and a second well of a second conductivity type, a first MOS transistor of the first conductivity type and a second MOS transistor of the second conductivity type. The first MOS transistor is disposed on the second well, including a gate structure on the second well and a strained layer of the first conductivity type in an opening in the second well beside the gate structure. The difference between the cell parameter of a portion of the strained layer near the bottom of the opening and that of the substrate is less than the difference between the cell parameter of a portion of the strained layer apart from the bottom of the opening and that of the substrate. The second MOS transistor is disposed on the first well.
Abstract:
This invention relates to poly(trimethylene terephthalate)/poly(alpha-hydroxy acid) biconstituent filaments, methods for making the same and end uses thereof.
Abstract:
A dynamic pulse width modulation (PWM) amplifier comprising an input terminal, a dynamic PWM controller, a power stage, a low pass filter, and an output terminal. The input terminal receives an input signal. The dynamic PWM controller transforms the N-bit input signal sampled by a sampling frequency to a 1-bit PWM signal. The power stage receives and outputs thel-bit PWM signal. The low pass filter receives and outputs the 1-bit PWM signal. The 1-bit PWM signal is used to drive the power stage and the low pass filter. The output terminal outputs the 1-bit PWM signal. The dynamic PWM amplifier is characterized in the dynamic PWM amplifier uses a register array to store the input signal processed immediately in each frame, regroups the input signal, and outputs the regroup signal, so that the 1-bit PWM signal is changed according to the input signal.