I am using a dataset from Google which contains 1,27,000 data points on simulated concentrations of the atmosphere of exoplanets which can sustain life. So, the output label of all these data points is 1 i.e, probability of life existing there is 1. If I train my neural network on this data, and test it on data points with concentrations other than these, can I expect to get probability values at the output? Asking because the model knows no false labelled value.
Yes, you can and the answer is One-Class Classification. A well-written resource to understand is this.