I was working with data consisting of parameters from a Java program running on different platforms. Unfortunately, this data has been lost. These parameters were used to determine whether errors would occur during the program's execution.
Initially, I used a logistic regression (LR) algorithm to predict if errors would occur based on the given parameters. Then, for the errored data, I employed a random forest algorithm to determine the number of errors that would occur.
For the first algorithm, I utilized a different algorithm called "gapper," which I developed myself. Thanks to this, I achieved 100% accuracy. However, websites suggest that an accuracy of around 70% is expected.