The task involves developing a machine learning model trained on urine output trends, clinical parameters, medications, and fluid input of patients to predict their future urine output.
What machine learning model would be most suitable for predicting a patient's urine output, considering the complex relationships between clinical parameters and historical data, and the sequential nature of urine output trends?
All algorithms have inductive biases which determine what they will be good or bad at. Which model is good on a set of data is determined by the data, not the model. For instance, in classification, Support Vector Machine works wonders on linearly separable data. Whether or not the classes are linearly separable and/or by which features is born out by the data.
It's best to start with a bit of data exploration to see which features are important for prediction and one's favorite "simple" model (i.e. Decision Tree Regression) just to see if decent performance can be simply achieved.
Lastly, just because something has dependency in the time domain does not mean one has to jump to something complex like LSTM. Feel free to include values from previous time steps as features to the present prediction. If it's a predictor for future values K-nearest neighbor is as likely a candidate as the next.
This sounds like a tabular regression problem to me.
tl;dr: use h2o.ai's automl, it's plug and play quite easy. It will test various different algorithms on your problem and create an ensemble of the best ones (that will likely be better than any individual model).
More context: The likely best algorithm would be a decision tree ensemble since they are still state of the art in tabular ML problems. But honestly if your problem turns out to be very simple a linear model might even be sufficient. You just need to make sure you provide the right features (e.g. how much water they drink, time since last pee, body size etc...).
