the time series or recursion based, the weights and bias.. the inputs outputs targets,
but i am not able to grasp the subject practically.
for example : i have a data collected everyday for 3 years, i presume now i should be able to predict the value for the next year.
i have also tried to keep my understanding simple.. use 1 input (in my case temperature) 365*3 values.
but i am not understanding several factors.
my input will be just 1 yr values = 365.. or 3 yr values 365*3
what will be my target ?
what equation/formula should i use to calculate the hidden layer values ?