The perception feed forward Artificial Neural Network is designed for the

testing and training of practical data. Maximum practical sensitivity found in 94.55 % with 1% Pd doped SnO2 based thick film

gas sensor at 300°C. Gradient Descent

Back-propagation with adaptive learning rate network function, the maximum

sensitivity for tansin network transfer function was found to be 94.55% (300 ml

concentration) at 300°C compare to other transfer network function(purelin

& logsin). The minimum error found on Tansin network transfer function here

we used Gradient Descent Back-propagation with adaptive learning algorithm.The multilayer

perceptron feed forward ANN was design for the testing and training purpose

using Levenberg -Marquardt feed forward propagation. LEARNGDM is used as its

adaptation learning function and MSE is used as performance function.The

maximum sensitivity recorded for 1% Pd doping sensor was calculated as 94.55%

at 300°C.Levenberg -Marquardt feed forward propagation algorithm, the maximum

sensitivity in tansin network transfer function was found to be 94.54% at 300°C

compare to network other transfer function.

The maximum sensitivity recorded for 1% Pd doping sensor was 94.55% at 300°C.

When the sensitivity was tested by matlab software neural network tool through Gradient

Descent Backpropagation with adaptive learning rate algorithm network function

in purelin network transfer function, the

maximum sensitivity for network was found to be 94.54% at 300°C as compared to Levenberg -Marquardt

feed forward propagation algorithm in

tansin network transfer function, the maximum sensitivity for network was found

as 94.55 % at 300 °C . Levenberg -Marquardt feed forward propagation algorithm

is the most preferred technique in comparison to Gradient Descent

Backpropagation with adaptive learning rate algorithm.

eed forward propagation algorithm in

tansin network transfer function, the maximum sensitivity for network was found

as 94.55 % at 300 °C . Levenberg -Marquardt feed forward propagation algorithm

is the most preferred technique in comparison to Gradient Descent

Backpropagation with adaptive learning rate algorithm