I have a subject that is measured for a continuous quantity V at multiple subsequent points in time (t1, t2, t3 etc). At each measurement time, the quantity V is essentially ‘reset’ back to 0 and V is essentially ‘additive’ in nature, can I create a set of ‘pseudo-samples’ based on all the pairwise combinations of each measurement time? i.e. if the measurements at t1 and t2 were missed for some reason, the value of V at t3 would theoretically be equivalent to V1 + V2. In matrix form, my samples would exist along the sub-diagonal, so my questions is can all of the values in the upper triangle matrix be used as additional ‘samples’? The intent for these data will be to build a predictive model of the quantity V so I was thinking that adding these values in might be beneficial? The image below (attempts) to illustrate my question, Measurements are collected at V1, V2 etc. are the yellow ? valid samples as well (V1 + V2 + V3, V1 + V2, V2 + V3)

enter image description here




Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.