Hello,
I am fairly new to machine learning, but I trying to build a SVM model (and explore other models) for binary classification using impedance spectrum data, solution conductivity and pH. The impedance data is a 61x2 matrix while conductivity and pH will be a single value. Attached is an example of a single input of which I am hoping to classify as either 1 or 0.
Thanks,
Hunter
To use SVM and other classifiers, you need to ensure that data matrix has the same size (i.e., 62x2). You can examine the bag-of-word techniques, which focus on addressing issues similar to your problem (features with different size). Finally, by looking at your data, I did not understand your problem, is a regression problem (to predict conductivity based on Zreal and Zimag)?
Artur this is a single input X. I would have additional measurements with a different frequency spectrum and corresponding conductivity and pH measurements. So excluding the conductivity and pH measurements my total input matrix would be a 61 x 2 x n where n is my number of measurements taken.
I would recommend you to resize the matrix 61 x 2 to a vector of size 122 x 1 and append the conductivity value and pH value at the end of the vector. So that your final input vector size is: 124 x 1. Hope it helps.
- Badges
- Science topic