Currently I am working for the task of Emotion Recognition. I have extracted the audio features from the labelled data which is generated in ARFF format by using opensmile toolkit. But when I try to load this ARFF file in weka for classification, Weka states that it is unable to determine the structure of ARFF. I am very new to Machine Learning approaches. So, please guide me in solving this issue.
ARFF File:
@relation SMILEfeaturesLive
@attribute name string
@attribute frameTime numeric
@attribute pcm_intensity_sma_max numeric
@attribute pcm_intensity_sma_min numeric
@attribute pcm_intensity_sma_range numeric
@attribute pcm_intensity_sma_maxPos numeric
@attribute pcm_intensity_sma_minPos numeric
@attribute pcm_intensity_sma_amean numeric
@attribute pcm_intensity_sma_linregc1 numeric
@attribute pcm_intensity_sma_linregc2 numeric
@attribute pcm_intensity_sma_linregerrA numeric
@attribute pcm_intensity_sma_linregerrQ numeric
@attribute pcm_intensity_sma_stddev numeric
@attribute pcm_intensity_sma_skewness numeric
@attribute pcm_intensity_sma_kurtosis numeric
@attribute pcm_intensity_sma_quartile1 numeric
@attribute pcm_intensity_sma_quartile2 numeric
@attribute pcm_intensity_sma_quartile3 numeric
@attribute pcm_intensity_sma_iqr1-2 numeric
@attribute pcm_intensity_sma_iqr2-3 numeric
@attribute pcm_intensity_sma_iqr1-3 numeric
@data
'liveturn_0',0.012500,2.522825e-04,1.051732e-08,2.522719e-
04,6.200000e+01,2.270000e+02,1.216623e-05,-9.114493e-08,2.437965e-
05,1.574297e-05,1.073658e-09,3.352240e-
05,5.023878e+00,3.133391e+01,8.150823e-08,4.335529e-07,9.001472e-
06,3.520447e-07,8.567918e-06,8.919964e-06,5.250381e+00,2.064137e-
01,5.043967e+00,6.200000e+01,1.580000e+02,1.168339e+00,-4.039864e-
03,1.709681e+00,6.951298e-01,8.278338e-01,9.624170e-
01,1.801888e+00,6.487199e+00,4.485340e-01,7.672363e-
01,1.565122e+00,3.187023e-01,1.256349e+01,anger
Error generated in WEKA: Unable to determine structure as arff(Reason: Java.io.IOException: no valid attribute type or invalid enumeration, read Token[anger],line 993).
