unable to use my own data like .xlsx, csv, and arff

1 view (last 30 days)
First question:
[x,t] = house_dataset; net = feedforwardnet(10,'trainlm'); net = train(net,x,t); y = net(x) But, here i want to use my own data which is .xlsx format instead of house_dataset provided by matlab to train the feedforward neural network and then to predict. This .xlsx contains only numeric value.
here is my excel data. 12 row, 6 column data
44.38 44.57 44.05 44.28 10014700.00 44.28 44.60 44.76 44.26 44.45 9223800.00 44.45 43.21 44.72 43.21 44.43 17922900.00 44.43 43.45 43.52 42.91 43.38 9797500.00 43.38 43.99 43.99 43.42 43.53 12129100.00 43.53 44.30 44.36 43.76 44.11 12036200.00 44.11 43.65 44.42 43.50 44.37 13301000.00 44.37 43.56 43.82 43.30 43.65 10410700.00 43.65 44.37 44.54 43.35 43.53 12798400.00 43.53 44.12 44.59 44.00 44.42 13785800.00 44.42 43.07 44.16 42.86 43.93 18158200.00 43.93 43.18 43.26 42.76 42.96 9791100.00 42.96
I read it like num = xlsread('myExample.xlsx');
this is the result num =
1.0e+07 *
0.0000 0.0000 0.0000 0.0000 1.0015 0.0000
0.0000 0.0000 0.0000 0.0000 0.9224 0.0000
0.0000 0.0000 0.0000 0.0000 1.7923 0.0000
0.0000 0.0000 0.0000 0.0000 0.9798 0.0000
0.0000 0.0000 0.0000 0.0000 1.2129 0.0000
0.0000 0.0000 0.0000 0.0000 1.2036 0.0000
0.0000 0.0000 0.0000 0.0000 1.3301 0.0000
0.0000 0.0000 0.0000 0.0000 1.0411 0.0000
0.0000 0.0000 0.0000 0.0000 1.2798 0.0000
0.0000 0.0000 0.0000 0.0000 1.3786 0.0000
0.0000 0.0000 0.0000 0.0000 1.8158 0.0000
0.0000 0.0000 0.0000 0.0000 0.9791 0.0000
and also Min max range 42.7600 18158200 1.8158e+07
and i assigned it to [x, t] = num
but the result is "Too many output argument"
Second question:
My data has 7 columns and 1500 rows including column name and one row of date data. how can i import this .xlsx data and predict new data with the next date for which the data is to be predicted for. Please help me

Answers (1)

Perry
Perry on 18 Mar 2015
Edited: Perry on 19 Mar 2015
2 ways to import
go to 'file' then 'import' and you can do it there or use
num = xlsread( filename , sheet , xlRange , 'basic' ) (but data needs to be in your working directory otherwise you need to specify the path where to find it c:....
on the dataset and prediction i have no idea, but instead of working with neural networks i would first suggest you consider pcr or pls (pricipial component regression)
trialversion: eigenvector research: pls toolbox-no code needed, its a gui or u check out www.models.life.ku.dk its a collection of code for multivariate dataanalysis
u read your trainingdata into inputs
houseInputs=num;
and you read the data you wanne predict/classify to
houseTargest=....; % houstargets is just one row with 506 entries
then u start the NN gui with nnstart and follow the instruction and play around be aware of the fact that you cant be a master of stuff when u spend less time as possible on a subject, what i mean is question your results with common sense and try to reproduce and understand the tutorial
http://www.mathworks.com/help/nnet/gs/fit-data-with-a-neural-network.html
hope it helps....
  1 Comment
Andualem alemu
Andualem alemu on 18 Mar 2015
Edited: Andualem alemu on 18 Mar 2015
tnx. now I have updated my question with the data i am using. would you please help me on this one

Sign in to comment.

Categories

Find more on Sequence and Numeric Feature Data Workflows in Help Center and File Exchange

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!