When optimizing the performance of neural network-based chatbots, determining the optimizer is one of the most important aspects. Optimizers primarily control the adjustment of model parameters such as weight and bias to minimize a loss function during training. Adaptive optimizers such as ADAM have become a standard choice and are widely used for their invariant parameter updates' magnitudes concerning gradient scale variations, but often pose generalization problems. Alternatively, Stochastic Gradient Descent (SGD) with Momentum and the extension of ADAM, the ADAMW, offers several advantages. This study aims to compare and examine the effects of these optimizers on the chatbot CST dataset. The effectiveness of each optimizer is evaluat

In this work, the possibility of a multiwavelength mode-locked fiber laser generation based on Four-Wave Mixing (FWM) induced by Fe2O3-SiO2 nanocomposite material is investigated for the first time. A multiwavelength mode-locked pulses fiber laser are generated from Ytterbium–doped fiber laser (YDFL) due to the combined action of high nonlinear absorption and high refractive coefficients of Fe2O3-SiO2 nanocomposite incorporated inside YDFL ring cavity. Up to more than 20 lasing lines in the 1040–1070 nm band with an equally lines separation of ~0.6 nm have been observed by just simple variation of passive modulation of the state of the polarization and the pump power altogether. Moreover, a passively mode-locked operation of YDFL laser

In this paper, we deal with the problem of general matching of two images one of them has experienced geometrical transformations, to find the correspondence between two images. We develop the invariant moments for traditional techniques (moments of inertia) with new approach to enhance the performance for these methods. We test various projections directional moments, to extract the difference between Block Distance Moment (BDM) and evaluate their reliability. Three adaptive strategies are shown for projections directional moments, that are raster (vertical and horizontal) projection, Fan-Bean projection and new projection procedure that is the square projection method. Our paper started with the description of a new algorithm that is low