For financial time series, the generation of error bars on the point of prediction is important in order to estimate the corresponding risk. In recent years, optimization techniques-driven artificial intelligence has been used to make time series approaches more systematic and improve forecasting performance. This paper presents a local linear radial basis functional neural network (LLRBFNN) model for classifying finance data from Yahoo Inc. The LLRBFNN model is learned by using the hybrid technique of backpropagation and recursive least square algorithm. The LLRBFNN model uses a local linear model in between the hidden layer and the output layer in contrast to the weights connected from hidden layer to output layer in typical neural network models. The obtained prediction result is compared with multilayer perceptron and radial basis functional neural network with the parameters being trained by gradient descent learning method. The proposed technique provides a lower mean squared error and thus can be considered as superior to other models. The technique is also tested on linear data, i.e., diabetic data, to confirm the validity of the result obtained from the experiment. Keywords Local linear radial basis functional neural network (LLRBFNN) Á Radial basis functional neural network (RBFNN) Á Multilayer perceptron (MLP) Á Recursive least square (RLS) Á Mean squared error (MSE)

No abstract

Computerization has been a need in recent years. As a result, massive amounts of digital data have been accumulated in practically every industry. Data mining approaches have arisen to serve with this rich data and sparse informations. Data-mining is a process that uses computers to extract and examine hidden patterns in data. Users may understand and benefit from this data mining process of knowledge discovery. According to the value associated with each property, different weights are allocated to those attributes. Weights assigned by the doctors are taken into consideration in this investigation. The first technique given is multiclass weighted associative classification with confidence-based rule ranking, while the second strategy is enhanced, genetic-based rule selection presented as a final step is a distributed multiclass weighted associative classifier. The distributed weighted associative classification is utilized as a consequence of the findings that lower the cost of communication while maintaining the benefit of the centralized approach. The multi class classification's accuracy % rises while using weighted associative classification.

A diformylphenol Schiff base of triazole‐amine, 4‐methyl‐2, 6‐bis‐[(1H‐[1,2,4] triazol‐3‐ylimino)‐methyl]‐phenol (PTR) has been synthesised and characterised by spectroscopic data (ESI‐MS, UV‐vis, FT‐IR, NMR spectra). The probe, PTR, emits at 610 nm upon excitation at 380 nm and the emission is strongly intensified on interacting with Zn2+ ion in DMSO‐water (99:1, v/v; HEPES buffer, pH, 7.2) solution even in the presence of 15 other cations (Cu2+, Mn2+, Co2+, Ni2+, Pd2+, Cd2+, Pb2+, Hg2+, Fe3+, Cr3+, Al3+, Na+, K+, Ba2+, Ca2+) and the emission band has been shifted to 530 nm. The limit of detection of Zn2+, 0.30 μM is much lower than WHO recommended value (76 μM). The binding constant (Kd) is 4.2585 × 104 M−1. Selective and sensitive chemosensing behaviour of PTR to Zn2+ has been explained by switching off ESIPT quenching of the keto‐enol tautomerisation of the probe along with the chelation enhancement of fluorescence (CHEF) by binding with Zn2+. The Job's plot and NMR titration have been extended to measure the 1:1 M composition [PTR + Zn2+]. The energy calculation by the DFT computation of keto and enol isomers supports easy tautomerisation and helps ESIPT quenching by proton transfer. Intracellular Zn2+ ions in living cells of HEK293 (PBS: phosphate buffer saline; pH, 7.2) have also been identified by the probe, PTR, using fluorescence microscopic imaging technique.