Statistical Data Analysis
****, a specialty auditing firm, is looking for a statistical analysis expert to provide training to a handful of staff and to consult on an occasional basis as needs arise.+33 Other Responses
we are a start up / prop trading firm. We need help to make statistics in order to develop automated ema setups vs. Volatility. We are not a university, do not waste our time please. We need an expert. A quantitative trader with excellent skills in math. If you are french it is even better.
Thanks+10 Other Responses
i want to get help in this , writing and explaining to me each step please.If you are expert in this let me know asap.
I have more than 22.000 genes for 58 breast cancer patients their genes before and after receiving treatment
to select genes modelling i used significance analysis and false discovery rate control the i ordered the q values
the next stage, i need to select the number of hidden neurons which is from 2:20
it is supposed to generate two loops the first one to select genes for modelling and the second to select hidden neuron number
i did the loop for selecting hidden neuron number but i could not do the first loop
which should be like that
for(t in 2:20) # t (gene) select genes for modelling
for(h in 2:20) #select hidden neuron number
#use nnet for t top genes and h hidden neurons
i used total accuracy to evaluate each model and roc curves.
I got 19 roc curves plots with box plot
my question s:
how can i choose the best and worst models from these. Should i use the box plot if yes how can i do that.
Also, how can i analyses the roc curves plots
also, how can i do sammon mapping for the best and worst model
the last question: how can i get the performance 3d plot by using the function persp. I know i should to use the gene number which it is supposed i obtain it from the loop but i could not do it
thanks+10 Other Responses